Add files

.gitattributes

@@ -1,35 +0,0 @@
-*.7z filter=lfs diff=lfs merge=lfs -text
-*.arrow filter=lfs diff=lfs merge=lfs -text
-*.bin filter=lfs diff=lfs merge=lfs -text
-*.bz2 filter=lfs diff=lfs merge=lfs -text
-*.ckpt filter=lfs diff=lfs merge=lfs -text
-*.ftz filter=lfs diff=lfs merge=lfs -text
-*.gz filter=lfs diff=lfs merge=lfs -text
-*.h5 filter=lfs diff=lfs merge=lfs -text
-*.joblib filter=lfs diff=lfs merge=lfs -text
-*.lfs.* filter=lfs diff=lfs merge=lfs -text
-*.mlmodel filter=lfs diff=lfs merge=lfs -text
-*.model filter=lfs diff=lfs merge=lfs -text
-*.msgpack filter=lfs diff=lfs merge=lfs -text
-*.npy filter=lfs diff=lfs merge=lfs -text
-*.npz filter=lfs diff=lfs merge=lfs -text
-*.onnx filter=lfs diff=lfs merge=lfs -text
-*.ot filter=lfs diff=lfs merge=lfs -text
-*.parquet filter=lfs diff=lfs merge=lfs -text
-*.pb filter=lfs diff=lfs merge=lfs -text
-*.pickle filter=lfs diff=lfs merge=lfs -text
-*.pkl filter=lfs diff=lfs merge=lfs -text
-*.pt filter=lfs diff=lfs merge=lfs -text
-*.pth filter=lfs diff=lfs merge=lfs -text
-*.rar filter=lfs diff=lfs merge=lfs -text
-*.safetensors filter=lfs diff=lfs merge=lfs -text
-saved_model/**/* filter=lfs diff=lfs merge=lfs -text
-*.tar.* filter=lfs diff=lfs merge=lfs -text
-*.tar filter=lfs diff=lfs merge=lfs -text
-*.tflite filter=lfs diff=lfs merge=lfs -text
-*.tgz filter=lfs diff=lfs merge=lfs -text
-*.wasm filter=lfs diff=lfs merge=lfs -text
-*.xz filter=lfs diff=lfs merge=lfs -text
-*.zip filter=lfs diff=lfs merge=lfs -text
-*.zst filter=lfs diff=lfs merge=lfs -text
-*tfevents* filter=lfs diff=lfs merge=lfs -text

.gitignore

@@ -0,0 +1,9 @@
+# Byte-compiled / optimized / DLL files
+__pycache__
+*.egg-info
+*.py[cod]
+*$py.class
+
+# Temporary data
+.DS_Store
+._*

README.md

@@ -1,12 +1,12 @@
 ---
 title: UniPixel
-emoji: 📉
-colorFrom: red
-colorTo: purple
+emoji: 🔮
+colorFrom: purple
+colorTo: yellow
 sdk: gradio
 sdk_version: 5.48.0
 app_file: app.py
-pinned: false
----
-
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+pinned: true
+license: bsd-3-clause
+short_description: Unified Object Referring and Segmentation for Pixel-Level Visual Reasoning
+---

app.py

@@ -0,0 +1,435 @@
+# Copyright (c) 2025 Ye Liu. Licensed under the BSD-3-Clause license.
+
+import os
+import re
+import uuid
+from functools import partial
+
+import gradio as gr
+import imageio.v3 as iio
+import spaces
+import torch
+import torch.nn.functional as F
+import torchvision.transforms.functional as T
+from PIL import Image
+
+from unipixel.constants import MEM_TOKEN, SEG_TOKEN
+from unipixel.dataset.utils import process_vision_info
+from unipixel.model.builder import build_model
+from unipixel.utils.io import load_image, load_video
+from unipixel.utils.transforms import get_sam2_transform
+from unipixel.utils.visualizer import draw_mask, sample_color
+
+PATH = os.path.abspath(os.path.dirname(os.path.realpath(__file__)))
+
+MODEL = 'PolyU-ChenLab/UniPixel-3B'
+
+TITLE = 'UniPixel: Unified Object Referring and Segmentation for Pixel-Level Visual Reasoning'
+
+HEADER = """
+<p align="center" style="margin: 1em 0 2em;"><img width="280" src="https://raw.githubusercontent.com/PolyU-ChenLab/UniPixel/refs/heads/main/.github/logo.png"></p>
+<h3 align="center">Unified Object Referring and Segmentation for Pixel-Level Visual Reasoning</h3>
+<div style="display: flex; justify-content: center; gap: 5px;">
+    <a href="https://arxiv.org/abs/2509.18094" target="_blank"><img src="https://img.shields.io/badge/arXiv-2509.18094-red"></a>
+    <a href="https://polyu-chenlab.github.io/unipixel/" target="_blank"><img src="https://img.shields.io/badge/Project-Page-brightgreen"></a>
+    <a href="https://huggingface.co/collections/PolyU-ChenLab/unipixel-68cf7137013455e5b15962e8" target="_blank"><img src="https://img.shields.io/badge/%F0%9F%A4%97%20Hugging%20Face-Model-blue"></a>
+    <a href="https://huggingface.co/datasets/PolyU-ChenLab/UniPixel-SFT-1M" target="_blank"><img src="https://img.shields.io/badge/%F0%9F%A4%97%20Hugging%20Face-Dataset-orange"></a>
+    <a href="https://github.com/PolyU-ChenLab/UniPixel/blob/main/README.md" target="_blank"><img src="https://img.shields.io/badge/License-BSD--3--Clause-purple"></a>
+    <a href="https://github.com/PolyU-ChenLab/UniPixel" target="_blank"><img src="https://img.shields.io/github/stars/PolyU-ChenLab/UniPixel"></a>
+</div>
+<p style="margin-top: 1em;">UniPixel is a unified MLLM for pixel-level vision-language understanding. It flexibly supports a variety of fine-grained tasks, including image/video segmentation, regional understanding, and a novel PixelQA task that jointly requires object-centric referring, segmentation, and question-answering in videos. Please open an <a href="https://github.com/PolyU-ChenLab/UniPixel/issues/new" target="_blank">issue</a> if you meet any problems.</p>
+"""
+
+# https://github.com/gradio-app/gradio/pull/10552
+JS = """
+function init() {
+    if (window.innerWidth >= 1536) {
+        document.querySelector('main').style.maxWidth = '1536px'
+    }
+}
+"""
+
+model, processor = build_model(MODEL)
+device = next(model.parameters()).device
+
+sam2_transform = get_sam2_transform(model.config.sam2_image_size)
+
+colors = sample_color()
+color_map = {f'Target {i + 1}': f'#{int(c[0]):02x}{int(c[1]):02x}{int(c[2]):02x}' for i, c in enumerate(colors * 255)}
+color_map_light = {
+    f'Target {i + 1}': f'#{int(c[0] * 127.5 + 127.5):02x}{int(c[1] * 127.5 + 127.5):02x}{int(c[2] * 127.5 + 127.5):02x}'
+    for i, c in enumerate(colors)
+}
+
+
+def enable_btns():
+    return (gr.Button(interactive=True), ) * 4
+
+
+def disable_btns():
+    return (gr.Button(interactive=False), ) * 4
+
+
+def reset_seg():
+    return 16, gr.Button(interactive=False)
+
+
+def reset_reg():
+    return 1, gr.Button(interactive=False)
+
+
+def update_region(blob):
+    if blob['background'] is None or not blob['layers'][0].any():
+        return
+
+    region = blob['background'].copy()
+    region[blob['layers'][0][:, :, -1] == 0] = [0, 0, 0, 0]
+
+    return region
+
+
+def update_video(video, prompt_idx):
+    if video is None:
+        return
+
+    _, images = load_video(video, sample_frames=16)
+    path = images[prompt_idx - 1]
+
+    return path
+
+
+@spaces.GPU
+def infer_seg(media, query, sample_frames=16, media_type=None):
+    if not media:
+        gr.Warning('Please upload an image or a video.')
+        return None, None, None
+
+    if not query:
+        gr.Warning('Please provide a text prompt.')
+        return None, None, None
+
+    if any(media.endswith(k) for k in ('jpg', 'png')):
+        frames, images = load_image(media), [media]
+    else:
+        frames, images = load_video(media, sample_frames=sample_frames)
+
+    messages = [{
+        'role':
+        'user',
+        'content': [{
+            'type': 'video',
+            'video': images,
+            'min_pixels': 128 * 28 * 28,
+            'max_pixels': 256 * 28 * 28 * int(sample_frames / len(images))
+        }, {
+            'type': 'text',
+            'text': query
+        }]
+    }]
+
+    text = processor.apply_chat_template(messages, add_generation_prompt=True)
+
+    images, videos, kwargs = process_vision_info(messages, return_video_kwargs=True)
+
+    data = processor(text=[text], images=images, videos=videos, return_tensors='pt', **kwargs)
+
+    data['frames'] = [sam2_transform(frames).to(model.sam2.dtype)]
+    data['frame_size'] = [frames.shape[1:3]]
+
+    output_ids = model.generate(
+        **data.to(device),
+        do_sample=False,
+        temperature=None,
+        top_k=None,
+        top_p=None,
+        repetition_penalty=None,
+        max_new_tokens=512)
+
+    assert data.input_ids.size(0) == output_ids.size(0) == 1
+    output_ids = output_ids[0, data.input_ids.size(1):]
+
+    if output_ids[-1] == processor.tokenizer.eos_token_id:
+        output_ids = output_ids[:-1]
+
+    response = processor.decode(output_ids, clean_up_tokenization_spaces=False)
+    response = response.replace(f' {SEG_TOKEN}', SEG_TOKEN).replace(f'{SEG_TOKEN} ', SEG_TOKEN)
+
+    entities = []
+    for i, m in enumerate(re.finditer(re.escape(SEG_TOKEN), response)):
+        entities.append(dict(entity=f'Target {i + 1}', start=m.start(), end=m.end()))
+
+    answer = dict(text=response, entities=entities)
+
+    imgs = draw_mask(frames, model.seg, colors=colors)
+
+    path = f"/tmp/{uuid.uuid4().hex}.{'gif' if len(imgs) > 1 else 'png'}"
+    iio.imwrite(path, imgs, duration=100, loop=0)
+
+    if media_type == 'image':
+        if len(model.seg) >= 1:
+            masks = media, [(m[0, 0].numpy(), f'Target {i + 1}') for i, m in enumerate(model.seg)]
+        else:
+            masks = None
+    else:
+        masks = path
+
+    return answer, masks, path
+
+
+infer_seg_image = partial(infer_seg, media_type='image')
+infer_seg_video = partial(infer_seg, media_type='video')
+
+
+@spaces.GPU
+def infer_reg(blob, query, prompt_idx=1, video=None):
+    if blob['background'] is None:
+        gr.Warning('Please upload an image or a video.')
+        return
+
+    if not blob['layers'][0].any():
+        gr.Warning('Please provide a mask prompt.')
+        return
+
+    if not query:
+        gr.Warning('Please provide a text prompt.')
+        return
+
+    if video is None:
+        frames = torch.from_numpy(blob['background'][:, :, :3]).unsqueeze(0)
+        images = [Image.fromarray(blob['background'], mode='RGBA')]
+    else:
+        frames, images = load_video(video, sample_frames=16)
+
+    frame_size = frames.shape[1:3]
+
+    mask = torch.from_numpy(blob['layers'][0][:, :, -1]).unsqueeze(0) > 0
+
+    refer_mask = torch.zeros(frames.size(0), 1, *frame_size)
+    refer_mask[prompt_idx - 1] = mask
+
+    if refer_mask.size(0) % 2 != 0:
+        refer_mask = torch.cat((refer_mask, refer_mask[-1, None]))
+    refer_mask = refer_mask.flatten(1)
+    refer_mask = F.max_pool1d(refer_mask.transpose(-1, -2), kernel_size=2, stride=2).transpose(-1, -2)
+    refer_mask = refer_mask.view(-1, 1, *frame_size)
+
+    if video is None:
+        prefix = f'Here is an image with the following highlighted regions:\n[0]: <{prompt_idx}> {MEM_TOKEN}\n'
+    else:
+        prefix = f'Here is a video with {len(images)} frames denoted as <1> to <{len(images)}>. The highlighted regions are as follows:\n[0]: <{prompt_idx}>-<{prompt_idx + 1}> {MEM_TOKEN}\n'
+
+    messages = [{
+        'role':
+        'user',
+        'content': [{
+            'type': 'video',
+            'video': images,
+            'min_pixels': 128 * 28 * 28,
+            'max_pixels': 256 * 28 * 28 * int(16 / len(images))
+        }, {
+            'type': 'text',
+            'text': prefix + query
+        }]
+    }]
+
+    text = processor.apply_chat_template(messages, add_generation_prompt=True)
+
+    images, videos, kwargs = process_vision_info(messages, return_video_kwargs=True)
+
+    data = processor(text=[text], images=images, videos=videos, return_tensors='pt', **kwargs)
+
+    refer_mask = T.resize(refer_mask, (data['video_grid_thw'][0][1] * 14, data['video_grid_thw'][0][2] * 14))
+    refer_mask = F.max_pool2d(refer_mask, kernel_size=28, stride=28)
+    refer_mask = refer_mask > 0
+
+    data['frames'] = [sam2_transform(frames).to(model.sam2.dtype)]
+    data['frame_size'] = [frames.shape[1:3]]
+    data['refer_mask'] = [refer_mask]
+
+    output_ids = model.generate(
+        **data.to(device),
+        do_sample=False,
+        temperature=None,
+        top_k=None,
+        top_p=None,
+        repetition_penalty=None,
+        max_new_tokens=512)
+
+    assert data.input_ids.size(0) == output_ids.size(0) == 1
+    output_ids = output_ids[0, data.input_ids.size(1):]
+
+    if output_ids[-1] == processor.tokenizer.eos_token_id:
+        output_ids = output_ids[:-1]
+
+    response = processor.decode(output_ids, clean_up_tokenization_spaces=False)
+    response = response.replace(' [0]', '[0]').replace('[0] ', '[0]').replace('[0]', '<REGION>')
+
+    entities = []
+    for m in re.finditer(re.escape('<REGION>'), response):
+        entities.append(dict(entity='region', start=m.start(), end=m.end(), color="#f85050"))
+
+    answer = dict(text=response, entities=entities)
+
+    return answer
+
+
+def build_demo():
+    with gr.Blocks(title=TITLE, js=JS) as demo:
+        gr.HTML(HEADER)
+
+        with gr.Tab('Image Segmentation'):
+            download_btn_1 = gr.DownloadButton(label='📦 Download', interactive=False, render=False)
+            msk_1 = gr.AnnotatedImage(label='Segmentation Results', color_map=color_map, render=False)
+            ans_1 = gr.HighlightedText(
+                label='Model Response', color_map=color_map_light, show_inline_category=False, render=False)
+
+            with gr.Row():
+                with gr.Column():
+                    media_1 = gr.Image(type='filepath')
+
+                    sample_frames_1 = gr.Slider(1, 32, value=16, step=1, visible=False)
+
+                    query_1 = gr.Textbox(label='Text Prompt', placeholder='Please segment the...')
+
+                    with gr.Row():
+                        random_btn_1 = gr.Button(value='🔮 Random', visible=False)
+
+                        reset_btn_1 = gr.ClearButton([media_1, query_1, msk_1, ans_1], value='🗑️ Reset')
+                        reset_btn_1.click(reset_seg, None, [sample_frames_1, download_btn_1])
+
+                        download_btn_1.render()
+
+                        submit_btn_1 = gr.Button(value='🚀 Submit', variant='primary')
+                with gr.Column():
+                    msk_1.render()
+                    ans_1.render()
+
+            ctx_1 = submit_btn_1.click(disable_btns, None, [random_btn_1, reset_btn_1, download_btn_1, submit_btn_1])
+            ctx_1 = ctx_1.then(infer_seg_image, [media_1, query_1, sample_frames_1], [ans_1, msk_1, download_btn_1])
+            ctx_1.then(enable_btns, None, [random_btn_1, reset_btn_1, download_btn_1, submit_btn_1])
+
+        with gr.Tab('Video Segmentation'):
+            download_btn_2 = gr.DownloadButton(label='📦 Download', interactive=False, render=False)
+            msk_2 = gr.Image(label='Segmentation Results', render=False)
+            ans_2 = gr.HighlightedText(
+                label='Model Response', color_map=color_map_light, show_inline_category=False, render=False)
+
+            with gr.Row():
+                with gr.Column():
+                    media_2 = gr.Video()
+
+                    with gr.Accordion(label='Hyperparameters', open=False):
+                        sample_frames_2 = gr.Slider(
+                            1,
+                            32,
+                            value=16,
+                            step=1,
+                            interactive=True,
+                            label='Sample Frames',
+                            info='The number of frames to sample from a video (Default: 16)')
+
+                    query_2 = gr.Textbox(label='Text Prompt', placeholder='Please segment the...')
+
+                    with gr.Row():
+                        random_btn_2 = gr.Button(value='🔮 Random', visible=False)
+
+                        reset_btn_2 = gr.ClearButton([media_2, query_2, msk_2, ans_2], value='🗑️ Reset')
+                        reset_btn_2.click(reset_seg, None, [sample_frames_2, download_btn_2])
+
+                        download_btn_2.render()
+
+                        submit_btn_2 = gr.Button(value='🚀 Submit', variant='primary')
+                with gr.Column():
+                    msk_2.render()
+                    ans_2.render()
+
+            ctx_2 = submit_btn_2.click(disable_btns, None, [random_btn_2, reset_btn_2, download_btn_2, submit_btn_2])
+            ctx_2 = ctx_2.then(infer_seg_video, [media_2, query_2, sample_frames_2], [ans_2, msk_2, download_btn_2])
+            ctx_2.then(enable_btns, None, [random_btn_2, reset_btn_2, download_btn_2, submit_btn_2])
+
+        with gr.Tab('Image Regional Understanding'):
+            download_btn_3 = gr.DownloadButton(visible=False)
+            msk_3 = gr.Image(label='Highlighted Region', render=False)
+            ans_3 = gr.HighlightedText(label='Model Response', show_inline_category=False, render=False)
+
+            with gr.Row():
+                with gr.Column():
+                    media_3 = gr.ImageEditor(
+                        label='Image & Mask Prompt',
+                        brush=gr.Brush(colors=["#ff000080"], color_mode='fixed'),
+                        transforms=None,
+                        layers=False)
+                    media_3.change(update_region, media_3, msk_3)
+
+                    prompt_frame_index_3 = gr.Slider(1, 16, value=1, step=1, visible=False)
+
+                    query_3 = gr.Textbox(label='Text Prompt', placeholder='Please describe the highlighted region...')
+
+                    with gr.Row():
+                        random_btn_3 = gr.Button(value='🔮 Random', visible=False)
+
+                        reset_btn_3 = gr.ClearButton([media_3, query_3, msk_3, ans_3], value='🗑️ Reset')
+                        reset_btn_3.click(reset_reg, None, [prompt_frame_index_3, download_btn_3])
+
+                        submit_btn_3 = gr.Button(value='🚀 Submit', variant='primary')
+                with gr.Column():
+                    msk_3.render()
+                    ans_3.render()
+
+            ctx_3 = submit_btn_3.click(disable_btns, None, [random_btn_3, reset_btn_3, download_btn_3, submit_btn_3])
+            ctx_3 = ctx_3.then(infer_reg, [media_3, query_3, prompt_frame_index_3], ans_3)
+            ctx_3.then(enable_btns, None, [random_btn_3, reset_btn_3, download_btn_3, submit_btn_3])
+
+        with gr.Tab('Video Regional Understanding'):
+            download_btn_4 = gr.DownloadButton(visible=False)
+            prompt_frame_index_4 = gr.Slider(
+                1,
+                16,
+                value=1,
+                step=1,
+                interactive=True,
+                label='Prompt Frame Index',
+                info='The index of the frame that includes mask prompts (Default: 1)',
+                render=False)
+            msk_4 = gr.ImageEditor(
+                label='Mask Prompt',
+                brush=gr.Brush(colors=['#ff000080'], color_mode='fixed'),
+                transforms=None,
+                layers=False,
+                render=False)
+            ans_4 = gr.HighlightedText(label='Model Response', show_inline_category=False, render=False)
+
+            with gr.Row():
+                with gr.Column():
+                    media_4 = gr.Video()
+                    media_4.change(update_video, [media_4, prompt_frame_index_4], msk_4)
+
+                    with gr.Accordion(label='Hyperparameters', open=False):
+                        prompt_frame_index_4.render()
+                        prompt_frame_index_4.change(update_video, [media_4, prompt_frame_index_4], msk_4)
+
+                    query_4 = gr.Textbox(label='Text Prompt', placeholder='Please describe the highlighted region...')
+
+                    with gr.Row():
+                        random_btn_4 = gr.Button(value='🔮 Random', visible=False)
+
+                        reset_btn_4 = gr.ClearButton([media_4, query_4, msk_4, ans_4], value='🗑️ Reset')
+                        reset_btn_4.click(reset_reg, None, [prompt_frame_index_4, download_btn_4])
+
+                        submit_btn_4 = gr.Button(value='🚀 Submit', variant='primary')
+                with gr.Column():
+                    msk_4.render()
+                    ans_4.render()
+
+            ctx_4 = submit_btn_4.click(disable_btns, None, [random_btn_4, reset_btn_4, download_btn_4, submit_btn_4])
+            ctx_4 = ctx_4.then(infer_reg, [msk_4, query_4, prompt_frame_index_4, media_4], ans_4)
+            ctx_4.then(enable_btns, None, [random_btn_4, reset_btn_4, download_btn_4, submit_btn_4])
+
+    return demo
+
+
+if __name__ == '__main__':
+    demo = build_demo()
+
+    demo.queue()
+    demo.launch(server_name='0.0.0.0')

requirements.txt

@@ -0,0 +1,33 @@
+accelerate==1.9.0
+decord==0.6.0
+deepspeed==0.17.4
+gradio==5.48.0
+hydra-core==1.3.2
+imageio==2.37.0
+iopath==0.1.10
+matplotlib==3.10.5
+nncore==0.4.7
+numpy==2.1.2
+openai==1.99.1
+pandas==2.3.1
+peft==0.17.0
+pycocotools==2.0.10
+pydantic==2.11.7
+pysrt==1.1.2
+scikit-image==0.25.2
+scikit-learn==1.7.1
+sentencepiece==0.2.0
+spaces==0.42.1
+tensordict==0.9.1
+termplotlib==0.3.9
+transformers==4.53.3
+triton==3.3.1
+wandb==0.21.0
+
+# torch==2.7.1+cu128
+# torchvision==0.22.1+cu128
+
+# https://github.com/Dao-AILab/flash-attention/pull/1751
+# flash_attn==2.8.2
+
+# sam2 modified from https://github.com/facebookresearch/sam2/tree/722d1d15111c689908aeeb82d49a57780aac5153

sam2/__init__.py

@@ -0,0 +1,11 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.
+
+from hydra import initialize_config_module
+from hydra.core.global_hydra import GlobalHydra
+
+if not GlobalHydra.instance().is_initialized():
+    initialize_config_module("sam2", version_base="1.2")

sam2/automatic_mask_generator.py

@@ -0,0 +1,416 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.
+
+# Adapted from https://github.com/facebookresearch/segment-anything/blob/main/segment_anything/automatic_mask_generator.py
+from typing import Any, Dict, List, Optional, Tuple
+
+import numpy as np
+import torch
+from torchvision.ops.boxes import batched_nms, box_area  # type: ignore
+
+from sam2.modeling.sam2_base import SAM2Base
+from sam2.sam2_image_predictor import SAM2ImagePredictor
+from sam2.utils.amg import (MaskData, area_from_rle, batch_iterator, batched_mask_to_box, box_xyxy_to_xywh,
+                            build_all_layer_point_grids, calculate_stability_score, coco_encode_rle,
+                            generate_crop_boxes, is_box_near_crop_edge, mask_to_rle_pytorch, remove_small_regions,
+                            rle_to_mask, uncrop_boxes_xyxy, uncrop_masks, uncrop_points)
+
+
+class SAM2AutomaticMaskGenerator:
+
+    def __init__(
+        self,
+        model: SAM2Base,
+        points_per_side: Optional[int] = 32,
+        points_per_batch: int = 64,
+        pred_iou_thresh: float = 0.8,
+        stability_score_thresh: float = 0.95,
+        stability_score_offset: float = 1.0,
+        mask_threshold: float = 0.0,
+        box_nms_thresh: float = 0.7,
+        crop_n_layers: int = 0,
+        crop_nms_thresh: float = 0.7,
+        crop_overlap_ratio: float = 512 / 1500,
+        crop_n_points_downscale_factor: int = 1,
+        point_grids: Optional[List[np.ndarray]] = None,
+        min_mask_region_area: int = 0,
+        output_mode: str = "binary_mask",
+        use_m2m: bool = False,
+        multimask_output: bool = True,
+        **kwargs,
+    ) -> None:
+        """
+        Using a SAM 2 model, generates masks for the entire image.
+        Generates a grid of point prompts over the image, then filters
+        low quality and duplicate masks. The default settings are chosen
+        for SAM 2 with a HieraL backbone.
+
+        Arguments:
+          model (Sam): The SAM 2 model to use for mask prediction.
+          points_per_side (int or None): The number of points to be sampled
+            along one side of the image. The total number of points is
+            points_per_side**2. If None, 'point_grids' must provide explicit
+            point sampling.
+          points_per_batch (int): Sets the number of points run simultaneously
+            by the model. Higher numbers may be faster but use more GPU memory.
+          pred_iou_thresh (float): A filtering threshold in [0,1], using the
+            model's predicted mask quality.
+          stability_score_thresh (float): A filtering threshold in [0,1], using
+            the stability of the mask under changes to the cutoff used to binarize
+            the model's mask predictions.
+          stability_score_offset (float): The amount to shift the cutoff when
+            calculated the stability score.
+          mask_threshold (float): Threshold for binarizing the mask logits
+          box_nms_thresh (float): The box IoU cutoff used by non-maximal
+            suppression to filter duplicate masks.
+          crop_n_layers (int): If >0, mask prediction will be run again on
+            crops of the image. Sets the number of layers to run, where each
+            layer has 2**i_layer number of image crops.
+          crop_nms_thresh (float): The box IoU cutoff used by non-maximal
+            suppression to filter duplicate masks between different crops.
+          crop_overlap_ratio (float): Sets the degree to which crops overlap.
+            In the first crop layer, crops will overlap by this fraction of
+            the image length. Later layers with more crops scale down this overlap.
+          crop_n_points_downscale_factor (int): The number of points-per-side
+            sampled in layer n is scaled down by crop_n_points_downscale_factor**n.
+          point_grids (list(np.ndarray) or None): A list over explicit grids
+            of points used for sampling, normalized to [0,1]. The nth grid in the
+            list is used in the nth crop layer. Exclusive with points_per_side.
+          min_mask_region_area (int): If >0, postprocessing will be applied
+            to remove disconnected regions and holes in masks with area smaller
+            than min_mask_region_area. Requires opencv.
+          output_mode (str): The form masks are returned in. Can be 'binary_mask',
+            'uncompressed_rle', or 'coco_rle'. 'coco_rle' requires pycocotools.
+            For large resolutions, 'binary_mask' may consume large amounts of
+            memory.
+          use_m2m (bool): Whether to add a one step refinement using previous mask predictions.
+          multimask_output (bool): Whether to output multimask at each point of the grid.
+        """
+
+        assert (points_per_side is None) != (point_grids
+                                             is None), "Exactly one of points_per_side or point_grid must be provided."
+        if points_per_side is not None:
+            self.point_grids = build_all_layer_point_grids(
+                points_per_side,
+                crop_n_layers,
+                crop_n_points_downscale_factor,
+            )
+        elif point_grids is not None:
+            self.point_grids = point_grids
+        else:
+            raise ValueError("Can't have both points_per_side and point_grid be None.")
+
+        assert output_mode in [
+            "binary_mask",
+            "uncompressed_rle",
+            "coco_rle",
+        ], f"Unknown output_mode {output_mode}."
+        if output_mode == "coco_rle":
+            try:
+                from pycocotools import mask as mask_utils  # type: ignore  # noqa: F401
+            except ImportError as e:
+                print("Please install pycocotools")
+                raise e
+
+        self.predictor = SAM2ImagePredictor(
+            model,
+            max_hole_area=min_mask_region_area,
+            max_sprinkle_area=min_mask_region_area,
+        )
+        self.points_per_batch = points_per_batch
+        self.pred_iou_thresh = pred_iou_thresh
+        self.stability_score_thresh = stability_score_thresh
+        self.stability_score_offset = stability_score_offset
+        self.mask_threshold = mask_threshold
+        self.box_nms_thresh = box_nms_thresh
+        self.crop_n_layers = crop_n_layers
+        self.crop_nms_thresh = crop_nms_thresh
+        self.crop_overlap_ratio = crop_overlap_ratio
+        self.crop_n_points_downscale_factor = crop_n_points_downscale_factor
+        self.min_mask_region_area = min_mask_region_area
+        self.output_mode = output_mode
+        self.use_m2m = use_m2m
+        self.multimask_output = multimask_output
+
+    @classmethod
+    def from_pretrained(cls, model_id: str, **kwargs) -> "SAM2AutomaticMaskGenerator":
+        """
+        Load a pretrained model from the Hugging Face hub.
+
+        Arguments:
+          model_id (str): The Hugging Face repository ID.
+          **kwargs: Additional arguments to pass to the model constructor.
+
+        Returns:
+          (SAM2AutomaticMaskGenerator): The loaded model.
+        """
+        from sam2.build_sam import build_sam2_hf
+
+        sam_model = build_sam2_hf(model_id, **kwargs)
+        return cls(sam_model, **kwargs)
+
+    @torch.no_grad()
+    def generate(self, image: np.ndarray) -> List[Dict[str, Any]]:
+        """
+        Generates masks for the given image.
+
+        Arguments:
+          image (np.ndarray): The image to generate masks for, in HWC uint8 format.
+
+        Returns:
+           list(dict(str, any)): A list over records for masks. Each record is
+             a dict containing the following keys:
+               segmentation (dict(str, any) or np.ndarray): The mask. If
+                 output_mode='binary_mask', is an array of shape HW. Otherwise,
+                 is a dictionary containing the RLE.
+               bbox (list(float)): The box around the mask, in XYWH format.
+               area (int): The area in pixels of the mask.
+               predicted_iou (float): The model's own prediction of the mask's
+                 quality. This is filtered by the pred_iou_thresh parameter.
+               point_coords (list(list(float))): The point coordinates input
+                 to the model to generate this mask.
+               stability_score (float): A measure of the mask's quality. This
+                 is filtered on using the stability_score_thresh parameter.
+               crop_box (list(float)): The crop of the image used to generate
+                 the mask, given in XYWH format.
+        """
+
+        # Generate masks
+        mask_data = self._generate_masks(image)
+
+        # Encode masks
+        if self.output_mode == "coco_rle":
+            mask_data["segmentations"] = [coco_encode_rle(rle) for rle in mask_data["rles"]]
+        elif self.output_mode == "binary_mask":
+            mask_data["segmentations"] = [rle_to_mask(rle) for rle in mask_data["rles"]]
+        else:
+            mask_data["segmentations"] = mask_data["rles"]
+
+        # Write mask records
+        curr_anns = []
+        for idx in range(len(mask_data["segmentations"])):
+            ann = {
+                "segmentation": mask_data["segmentations"][idx],
+                "area": area_from_rle(mask_data["rles"][idx]),
+                "bbox": box_xyxy_to_xywh(mask_data["boxes"][idx]).tolist(),
+                "predicted_iou": mask_data["iou_preds"][idx].item(),
+                "point_coords": [mask_data["points"][idx].tolist()],
+                "stability_score": mask_data["stability_score"][idx].item(),
+                "crop_box": box_xyxy_to_xywh(mask_data["crop_boxes"][idx]).tolist(),
+            }
+            curr_anns.append(ann)
+
+        return curr_anns
+
+    def _generate_masks(self, image: np.ndarray) -> MaskData:
+        orig_size = image.shape[:2]
+        crop_boxes, layer_idxs = generate_crop_boxes(orig_size, self.crop_n_layers, self.crop_overlap_ratio)
+
+        # Iterate over image crops
+        data = MaskData()
+        for crop_box, layer_idx in zip(crop_boxes, layer_idxs):
+            crop_data = self._process_crop(image, crop_box, layer_idx, orig_size)
+            data.cat(crop_data)
+
+        # Remove duplicate masks between crops
+        if len(crop_boxes) > 1:
+            # Prefer masks from smaller crops
+            scores = 1 / box_area(data["crop_boxes"])
+            scores = scores.to(data["boxes"].device)
+            keep_by_nms = batched_nms(
+                data["boxes"].float(),
+                scores,
+                torch.zeros_like(data["boxes"][:, 0]),  # categories
+                iou_threshold=self.crop_nms_thresh,
+            )
+            data.filter(keep_by_nms)
+        data.to_numpy()
+        return data
+
+    def _process_crop(
+        self,
+        image: np.ndarray,
+        crop_box: List[int],
+        crop_layer_idx: int,
+        orig_size: Tuple[int, ...],
+    ) -> MaskData:
+        # Crop the image and calculate embeddings
+        x0, y0, x1, y1 = crop_box
+        cropped_im = image[y0:y1, x0:x1, :]
+        cropped_im_size = cropped_im.shape[:2]
+        self.predictor.set_image(cropped_im)
+
+        # Get points for this crop
+        points_scale = np.array(cropped_im_size)[None, ::-1]
+        points_for_image = self.point_grids[crop_layer_idx] * points_scale
+
+        # Generate masks for this crop in batches
+        data = MaskData()
+        for (points, ) in batch_iterator(self.points_per_batch, points_for_image):
+            batch_data = self._process_batch(points, cropped_im_size, crop_box, orig_size, normalize=True)
+            data.cat(batch_data)
+            del batch_data
+        self.predictor.reset_predictor()
+
+        # Remove duplicates within this crop.
+        keep_by_nms = batched_nms(
+            data["boxes"].float(),
+            data["iou_preds"],
+            torch.zeros_like(data["boxes"][:, 0]),  # categories
+            iou_threshold=self.box_nms_thresh,
+        )
+        data.filter(keep_by_nms)
+
+        # Return to the original image frame
+        data["boxes"] = uncrop_boxes_xyxy(data["boxes"], crop_box)
+        data["points"] = uncrop_points(data["points"], crop_box)
+        data["crop_boxes"] = torch.tensor([crop_box for _ in range(len(data["rles"]))])
+
+        return data
+
+    def _process_batch(
+        self,
+        points: np.ndarray,
+        im_size: Tuple[int, ...],
+        crop_box: List[int],
+        orig_size: Tuple[int, ...],
+        normalize=False,
+    ) -> MaskData:
+        orig_h, orig_w = orig_size
+
+        # Run model on this batch
+        points = torch.as_tensor(points, dtype=torch.float32, device=self.predictor.device)
+        in_points = self.predictor._transforms.transform_coords(points, normalize=normalize, orig_hw=im_size)
+        in_labels = torch.ones(in_points.shape[0], dtype=torch.int, device=in_points.device)
+        masks, iou_preds, low_res_masks = self.predictor._predict(
+            in_points[:, None, :],
+            in_labels[:, None],
+            multimask_output=self.multimask_output,
+            return_logits=True,
+        )
+
+        # Serialize predictions and store in MaskData
+        data = MaskData(
+            masks=masks.flatten(0, 1),
+            iou_preds=iou_preds.flatten(0, 1),
+            points=points.repeat_interleave(masks.shape[1], dim=0),
+            low_res_masks=low_res_masks.flatten(0, 1),
+        )
+        del masks
+
+        if not self.use_m2m:
+            # Filter by predicted IoU
+            if self.pred_iou_thresh > 0.0:
+                keep_mask = data["iou_preds"] > self.pred_iou_thresh
+                data.filter(keep_mask)
+
+            # Calculate and filter by stability score
+            data["stability_score"] = calculate_stability_score(data["masks"], self.mask_threshold,
+                                                                self.stability_score_offset)
+            if self.stability_score_thresh > 0.0:
+                keep_mask = data["stability_score"] >= self.stability_score_thresh
+                data.filter(keep_mask)
+        else:
+            # One step refinement using previous mask predictions
+            in_points = self.predictor._transforms.transform_coords(
+                data["points"], normalize=normalize, orig_hw=im_size)
+            labels = torch.ones(in_points.shape[0], dtype=torch.int, device=in_points.device)
+            masks, ious = self.refine_with_m2m(in_points, labels, data["low_res_masks"], self.points_per_batch)
+            data["masks"] = masks.squeeze(1)
+            data["iou_preds"] = ious.squeeze(1)
+
+            if self.pred_iou_thresh > 0.0:
+                keep_mask = data["iou_preds"] > self.pred_iou_thresh
+                data.filter(keep_mask)
+
+            data["stability_score"] = calculate_stability_score(data["masks"], self.mask_threshold,
+                                                                self.stability_score_offset)
+            if self.stability_score_thresh > 0.0:
+                keep_mask = data["stability_score"] >= self.stability_score_thresh
+                data.filter(keep_mask)
+
+        # Threshold masks and calculate boxes
+        data["masks"] = data["masks"] > self.mask_threshold
+        data["boxes"] = batched_mask_to_box(data["masks"])
+
+        # Filter boxes that touch crop boundaries
+        keep_mask = ~is_box_near_crop_edge(data["boxes"], crop_box, [0, 0, orig_w, orig_h])
+        if not torch.all(keep_mask):
+            data.filter(keep_mask)
+
+        # Compress to RLE
+        data["masks"] = uncrop_masks(data["masks"], crop_box, orig_h, orig_w)
+        data["rles"] = mask_to_rle_pytorch(data["masks"])
+        del data["masks"]
+
+        return data
+
+    @staticmethod
+    def postprocess_small_regions(mask_data: MaskData, min_area: int, nms_thresh: float) -> MaskData:
+        """
+        Removes small disconnected regions and holes in masks, then reruns
+        box NMS to remove any new duplicates.
+
+        Edits mask_data in place.
+
+        Requires open-cv as a dependency.
+        """
+        if len(mask_data["rles"]) == 0:
+            return mask_data
+
+        # Filter small disconnected regions and holes
+        new_masks = []
+        scores = []
+        for rle in mask_data["rles"]:
+            mask = rle_to_mask(rle)
+
+            mask, changed = remove_small_regions(mask, min_area, mode="holes")
+            unchanged = not changed
+            mask, changed = remove_small_regions(mask, min_area, mode="islands")
+            unchanged = unchanged and not changed
+
+            new_masks.append(torch.as_tensor(mask).unsqueeze(0))
+            # Give score=0 to changed masks and score=1 to unchanged masks
+            # so NMS will prefer ones that didn't need postprocessing
+            scores.append(float(unchanged))
+
+        # Recalculate boxes and remove any new duplicates
+        masks = torch.cat(new_masks, dim=0)
+        boxes = batched_mask_to_box(masks)
+        keep_by_nms = batched_nms(
+            boxes.float(),
+            torch.as_tensor(scores),
+            torch.zeros_like(boxes[:, 0]),  # categories
+            iou_threshold=nms_thresh,
+        )
+
+        # Only recalculate RLEs for masks that have changed
+        for i_mask in keep_by_nms:
+            if scores[i_mask] == 0.0:
+                mask_torch = masks[i_mask].unsqueeze(0)
+                mask_data["rles"][i_mask] = mask_to_rle_pytorch(mask_torch)[0]
+                mask_data["boxes"][i_mask] = boxes[i_mask]  # update res directly
+        mask_data.filter(keep_by_nms)
+
+        return mask_data
+
+    def refine_with_m2m(self, points, point_labels, low_res_masks, points_per_batch):
+        new_masks = []
+        new_iou_preds = []
+
+        for cur_points, cur_point_labels, low_res_mask in batch_iterator(points_per_batch, points, point_labels,
+                                                                         low_res_masks):
+            best_masks, best_iou_preds, _ = self.predictor._predict(
+                cur_points[:, None, :],
+                cur_point_labels[:, None],
+                mask_input=low_res_mask[:, None, :],
+                multimask_output=False,
+                return_logits=True,
+            )
+            new_masks.append(best_masks)
+            new_iou_preds.append(best_iou_preds)
+        masks = torch.cat(new_masks, dim=0)
+        return masks, torch.cat(new_iou_preds, dim=0)

sam2/build_sam.py

@@ -0,0 +1,172 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.
+
+import logging
+import os
+
+import torch
+from hydra import compose
+from hydra.utils import instantiate
+from omegaconf import OmegaConf
+
+import sam2
+
+# Check if the user is running Python from the parent directory of the sam2 repo
+# (i.e. the directory where this repo is cloned into) -- this is not supported since
+# it could shadow the sam2 package and cause issues.
+if os.path.isdir(os.path.join(sam2.__path__[0], "sam2")):
+    # If the user has "sam2/sam2" in their path, they are likey importing the repo itself
+    # as "sam2" rather than importing the "sam2" python package (i.e. "sam2/sam2" directory).
+    # This typically happens because the user is running Python from the parent directory
+    # that contains the sam2 repo they cloned.
+    raise RuntimeError("You're likely running Python from the parent directory of the sam2 repository "
+                       "(i.e. the directory where https://github.com/facebookresearch/sam2 is cloned into). "
+                       "This is not supported since the `sam2` Python package could be shadowed by the "
+                       "repository name (the repository is also named `sam2` and contains the Python package "
+                       "in `sam2/sam2`). Please run Python from another directory (e.g. from the repo dir "
+                       "rather than its parent dir, or from your home directory) after installing SAM 2.")
+
+HF_MODEL_ID_TO_FILENAMES = {
+    "facebook/sam2-hiera-tiny": (
+        "configs/sam2/sam2_hiera_t.yaml",
+        "sam2_hiera_tiny.pt",
+    ),
+    "facebook/sam2-hiera-small": (
+        "configs/sam2/sam2_hiera_s.yaml",
+        "sam2_hiera_small.pt",
+    ),
+    "facebook/sam2-hiera-base-plus": (
+        "configs/sam2/sam2_hiera_b+.yaml",
+        "sam2_hiera_base_plus.pt",
+    ),
+    "facebook/sam2-hiera-large": (
+        "configs/sam2/sam2_hiera_l.yaml",
+        "sam2_hiera_large.pt",
+    ),
+    "facebook/sam2.1-hiera-tiny": (
+        "configs/sam2.1/sam2.1_hiera_t.yaml",
+        "sam2.1_hiera_tiny.pt",
+    ),
+    "facebook/sam2.1-hiera-small": (
+        "configs/sam2.1/sam2.1_hiera_s.yaml",
+        "sam2.1_hiera_small.pt",
+    ),
+    "facebook/sam2.1-hiera-base-plus": (
+        "configs/sam2.1/sam2.1_hiera_b+.yaml",
+        "sam2.1_hiera_base_plus.pt",
+    ),
+    "facebook/sam2.1-hiera-large": (
+        "configs/sam2.1/sam2.1_hiera_l.yaml",
+        "sam2.1_hiera_large.pt",
+    ),
+}
+
+
+def build_sam2(
+    config_file,
+    ckpt_path=None,
+    device="cuda",
+    mode="eval",
+    hydra_overrides_extra=[],
+    apply_postprocessing=True,
+    **kwargs,
+):
+
+    if apply_postprocessing:
+        hydra_overrides_extra = hydra_overrides_extra.copy()
+        hydra_overrides_extra += [
+            # dynamically fall back to multi-mask if the single mask is not stable
+            "++model.sam_mask_decoder_extra_args.dynamic_multimask_via_stability=true",
+            "++model.sam_mask_decoder_extra_args.dynamic_multimask_stability_delta=0.05",
+            "++model.sam_mask_decoder_extra_args.dynamic_multimask_stability_thresh=0.98",
+        ]
+    # Read config and init model
+    cfg = compose(config_name=config_file, overrides=hydra_overrides_extra)
+    OmegaConf.resolve(cfg)
+    model = instantiate(cfg.model, _recursive_=True)
+    _load_checkpoint(model, ckpt_path)
+    model = model.to(device)
+    if mode == "eval":
+        model.eval()
+    return model
+
+
+def build_sam2_video_predictor(
+    config_file,
+    ckpt_path=None,
+    device="cuda",
+    mode="eval",
+    hydra_overrides_extra=[],
+    apply_postprocessing=True,
+    vos_optimized=False,
+    **kwargs,
+):
+    hydra_overrides = [
+        "++model._target_=sam2.sam2_video_predictor.SAM2VideoPredictor",
+    ]
+    if vos_optimized:
+        hydra_overrides = [
+            "++model._target_=sam2.sam2_video_predictor.SAM2VideoPredictorVOS",
+            "++model.compile_image_encoder=True",  # Let sam2_base handle this
+        ]
+
+    if apply_postprocessing:
+        hydra_overrides_extra = hydra_overrides_extra.copy()
+        hydra_overrides_extra += [
+            # dynamically fall back to multi-mask if the single mask is not stable
+            "++model.sam_mask_decoder_extra_args.dynamic_multimask_via_stability=true",
+            "++model.sam_mask_decoder_extra_args.dynamic_multimask_stability_delta=0.05",
+            "++model.sam_mask_decoder_extra_args.dynamic_multimask_stability_thresh=0.98",
+            # the sigmoid mask logits on interacted frames with clicks in the memory encoder so that the encoded masks are exactly as what users see from clicking
+            "++model.binarize_mask_from_pts_for_mem_enc=true",
+            # fill small holes in the low-res masks up to `fill_hole_area` (before resizing them to the original video resolution)
+            "++model.fill_hole_area=8",
+        ]
+    hydra_overrides.extend(hydra_overrides_extra)
+
+    # Read config and init model
+    cfg = compose(config_name=config_file, overrides=hydra_overrides)
+    OmegaConf.resolve(cfg)
+    model = instantiate(cfg.model, _recursive_=True)
+    _load_checkpoint(model, ckpt_path)
+    model = model.to(device)
+    if mode == "eval":
+        model.eval()
+    return model
+
+
+def _hf_download(model_id):
+    from huggingface_hub import hf_hub_download
+
+    config_name, checkpoint_name = HF_MODEL_ID_TO_FILENAMES[model_id]
+    ckpt_path = hf_hub_download(repo_id=model_id, filename=checkpoint_name)
+    return config_name, ckpt_path
+
+
+def build_sam2_hf(model_id, **kwargs):
+    config_name, ckpt_path = _hf_download(model_id)
+    return build_sam2(config_file=config_name, ckpt_path=ckpt_path, **kwargs)
+
+
+def build_sam2_video_predictor_hf(model_id, **kwargs):
+    config_name, ckpt_path = _hf_download(model_id)
+    return build_sam2_video_predictor(config_file=config_name, ckpt_path=ckpt_path, **kwargs)
+
+
+def _load_checkpoint(model, ckpt_path):
+    if ckpt_path is not None:
+        sd = torch.load(ckpt_path, map_location="cpu", weights_only=True)["model"]
+        # https://github.com/huggingface/transformers/issues/29554
+        sd['memory_encoder.fuser.layers.0.weight'] = sd.pop('memory_encoder.fuser.layers.0.gamma')
+        sd['memory_encoder.fuser.layers.1.weight'] = sd.pop('memory_encoder.fuser.layers.1.gamma')
+        missing_keys, unexpected_keys = model.load_state_dict(sd)
+        if missing_keys:
+            logging.error(missing_keys)
+            raise RuntimeError()
+        if unexpected_keys:
+            logging.error(unexpected_keys)
+            raise RuntimeError()
+        logging.info("Loaded checkpoint sucessfully")

sam2/configs/sam2.1/sam2.1_hiera_b+.yaml

@@ -0,0 +1,116 @@
+# @package _global_
+
+# Model
+model:
+  _target_: sam2.modeling.sam2_base.SAM2Base
+  image_encoder:
+    _target_: sam2.modeling.backbones.image_encoder.ImageEncoder
+    scalp: 1
+    trunk:
+      _target_: sam2.modeling.backbones.hieradet.Hiera
+      embed_dim: 112
+      num_heads: 2
+    neck:
+      _target_: sam2.modeling.backbones.image_encoder.FpnNeck
+      position_encoding:
+        _target_: sam2.modeling.position_encoding.PositionEmbeddingSine
+        num_pos_feats: 256
+        normalize: true
+        scale: null
+        temperature: 10000
+      d_model: 256
+      backbone_channel_list: [896, 448, 224, 112]
+      fpn_top_down_levels: [2, 3]  # output level 0 and 1 directly use the backbone features
+      fpn_interp_model: nearest
+
+  memory_attention:
+    _target_: sam2.modeling.memory_attention.MemoryAttention
+    d_model: 256
+    pos_enc_at_input: true
+    layer:
+      _target_: sam2.modeling.memory_attention.MemoryAttentionLayer
+      activation: relu
+      dim_feedforward: 2048
+      dropout: 0.1
+      pos_enc_at_attn: false
+      self_attention:
+        _target_: sam2.modeling.sam.transformer.RoPEAttention
+        rope_theta: 10000.0
+        feat_sizes: [64, 64]
+        embedding_dim: 256
+        num_heads: 1
+        downsample_rate: 1
+        dropout: 0.1
+      d_model: 256
+      pos_enc_at_cross_attn_keys: true
+      pos_enc_at_cross_attn_queries: false
+      cross_attention:
+        _target_: sam2.modeling.sam.transformer.RoPEAttention
+        rope_theta: 10000.0
+        feat_sizes: [64, 64]
+        rope_k_repeat: True
+        embedding_dim: 256
+        num_heads: 1
+        downsample_rate: 1
+        dropout: 0.1
+        kv_in_dim: 64
+    num_layers: 4
+
+  memory_encoder:
+      _target_: sam2.modeling.memory_encoder.MemoryEncoder
+      out_dim: 64
+      position_encoding:
+        _target_: sam2.modeling.position_encoding.PositionEmbeddingSine
+        num_pos_feats: 64
+        normalize: true
+        scale: null
+        temperature: 10000
+      mask_downsampler:
+        _target_: sam2.modeling.memory_encoder.MaskDownSampler
+        kernel_size: 3
+        stride: 2
+        padding: 1
+      fuser:
+        _target_: sam2.modeling.memory_encoder.Fuser
+        layer:
+          _target_: sam2.modeling.memory_encoder.CXBlock
+          dim: 256
+          kernel_size: 7
+          padding: 3
+          layer_scale_init_value: 1e-6
+          use_dwconv: True  # depth-wise convs
+        num_layers: 2
+
+  num_maskmem: 7
+  image_size: 1024
+  # apply scaled sigmoid on mask logits for memory encoder, and directly feed input mask as output mask
+  sigmoid_scale_for_mem_enc: 20.0
+  sigmoid_bias_for_mem_enc: -10.0
+  use_mask_input_as_output_without_sam: true
+  # Memory
+  directly_add_no_mem_embed: true
+  no_obj_embed_spatial: true
+  # use high-resolution feature map in the SAM mask decoder
+  use_high_res_features_in_sam: true
+  # output 3 masks on the first click on initial conditioning frames
+  multimask_output_in_sam: true
+  # SAM heads
+  iou_prediction_use_sigmoid: True
+  # cross-attend to object pointers from other frames (based on SAM output tokens) in the encoder
+  use_obj_ptrs_in_encoder: true
+  add_tpos_enc_to_obj_ptrs: true
+  proj_tpos_enc_in_obj_ptrs: true
+  use_signed_tpos_enc_to_obj_ptrs: true
+  only_obj_ptrs_in_the_past_for_eval: true
+  # object occlusion prediction
+  pred_obj_scores: true
+  pred_obj_scores_mlp: true
+  fixed_no_obj_ptr: true
+  # multimask tracking settings
+  multimask_output_for_tracking: true
+  use_multimask_token_for_obj_ptr: true
+  multimask_min_pt_num: 0
+  multimask_max_pt_num: 1
+  use_mlp_for_obj_ptr_proj: true
+  # Compilation flag
+  compile_image_encoder: False

sam2/configs/sam2.1/sam2.1_hiera_l.yaml

@@ -0,0 +1,120 @@
+# @package _global_
+
+# Model
+model:
+  _target_: sam2.modeling.sam2_base.SAM2Base
+  image_encoder:
+    _target_: sam2.modeling.backbones.image_encoder.ImageEncoder
+    scalp: 1
+    trunk:
+      _target_: sam2.modeling.backbones.hieradet.Hiera
+      embed_dim: 144
+      num_heads: 2
+      stages: [2, 6, 36, 4]
+      global_att_blocks: [23, 33, 43]
+      window_pos_embed_bkg_spatial_size: [7, 7]
+      window_spec: [8, 4, 16, 8]
+    neck:
+      _target_: sam2.modeling.backbones.image_encoder.FpnNeck
+      position_encoding:
+        _target_: sam2.modeling.position_encoding.PositionEmbeddingSine
+        num_pos_feats: 256
+        normalize: true
+        scale: null
+        temperature: 10000
+      d_model: 256
+      backbone_channel_list: [1152, 576, 288, 144]
+      fpn_top_down_levels: [2, 3]  # output level 0 and 1 directly use the backbone features
+      fpn_interp_model: nearest
+
+  memory_attention:
+    _target_: sam2.modeling.memory_attention.MemoryAttention
+    d_model: 256
+    pos_enc_at_input: true
+    layer:
+      _target_: sam2.modeling.memory_attention.MemoryAttentionLayer
+      activation: relu
+      dim_feedforward: 2048
+      dropout: 0.1
+      pos_enc_at_attn: false
+      self_attention:
+        _target_: sam2.modeling.sam.transformer.RoPEAttention
+        rope_theta: 10000.0
+        feat_sizes: [64, 64]
+        embedding_dim: 256
+        num_heads: 1
+        downsample_rate: 1
+        dropout: 0.1
+      d_model: 256
+      pos_enc_at_cross_attn_keys: true
+      pos_enc_at_cross_attn_queries: false
+      cross_attention:
+        _target_: sam2.modeling.sam.transformer.RoPEAttention
+        rope_theta: 10000.0
+        feat_sizes: [64, 64]
+        rope_k_repeat: True
+        embedding_dim: 256
+        num_heads: 1
+        downsample_rate: 1
+        dropout: 0.1
+        kv_in_dim: 64
+    num_layers: 4
+
+  memory_encoder:
+      _target_: sam2.modeling.memory_encoder.MemoryEncoder
+      out_dim: 64
+      position_encoding:
+        _target_: sam2.modeling.position_encoding.PositionEmbeddingSine
+        num_pos_feats: 64
+        normalize: true
+        scale: null
+        temperature: 10000
+      mask_downsampler:
+        _target_: sam2.modeling.memory_encoder.MaskDownSampler
+        kernel_size: 3
+        stride: 2
+        padding: 1
+      fuser:
+        _target_: sam2.modeling.memory_encoder.Fuser
+        layer:
+          _target_: sam2.modeling.memory_encoder.CXBlock
+          dim: 256
+          kernel_size: 7
+          padding: 3
+          layer_scale_init_value: 1e-6
+          use_dwconv: True  # depth-wise convs
+        num_layers: 2
+
+  num_maskmem: 7
+  image_size: 1024
+  # apply scaled sigmoid on mask logits for memory encoder, and directly feed input mask as output mask
+  sigmoid_scale_for_mem_enc: 20.0
+  sigmoid_bias_for_mem_enc: -10.0
+  use_mask_input_as_output_without_sam: true
+  # Memory
+  directly_add_no_mem_embed: true
+  no_obj_embed_spatial: true
+  # use high-resolution feature map in the SAM mask decoder
+  use_high_res_features_in_sam: true
+  # output 3 masks on the first click on initial conditioning frames
+  multimask_output_in_sam: true
+  # SAM heads
+  iou_prediction_use_sigmoid: True
+  # cross-attend to object pointers from other frames (based on SAM output tokens) in the encoder
+  use_obj_ptrs_in_encoder: true
+  add_tpos_enc_to_obj_ptrs: true
+  proj_tpos_enc_in_obj_ptrs: true
+  use_signed_tpos_enc_to_obj_ptrs: true
+  only_obj_ptrs_in_the_past_for_eval: true
+  # object occlusion prediction
+  pred_obj_scores: true
+  pred_obj_scores_mlp: true
+  fixed_no_obj_ptr: true
+  # multimask tracking settings
+  multimask_output_for_tracking: true
+  use_multimask_token_for_obj_ptr: true
+  multimask_min_pt_num: 0
+  multimask_max_pt_num: 1
+  use_mlp_for_obj_ptr_proj: true
+  # Compilation flag
+  compile_image_encoder: False

sam2/configs/sam2.1/sam2.1_hiera_s.yaml

@@ -0,0 +1,119 @@
+# @package _global_
+
+# Model
+model:
+  _target_: sam2.modeling.sam2_base.SAM2Base
+  image_encoder:
+    _target_: sam2.modeling.backbones.image_encoder.ImageEncoder
+    scalp: 1
+    trunk:
+      _target_: sam2.modeling.backbones.hieradet.Hiera
+      embed_dim: 96
+      num_heads: 1
+      stages: [1, 2, 11, 2]
+      global_att_blocks: [7, 10, 13]
+      window_pos_embed_bkg_spatial_size: [7, 7]
+    neck:
+      _target_: sam2.modeling.backbones.image_encoder.FpnNeck
+      position_encoding:
+        _target_: sam2.modeling.position_encoding.PositionEmbeddingSine
+        num_pos_feats: 256
+        normalize: true
+        scale: null
+        temperature: 10000
+      d_model: 256
+      backbone_channel_list: [768, 384, 192, 96]
+      fpn_top_down_levels: [2, 3]  # output level 0 and 1 directly use the backbone features
+      fpn_interp_model: nearest
+
+  memory_attention:
+    _target_: sam2.modeling.memory_attention.MemoryAttention
+    d_model: 256
+    pos_enc_at_input: true
+    layer:
+      _target_: sam2.modeling.memory_attention.MemoryAttentionLayer
+      activation: relu
+      dim_feedforward: 2048
+      dropout: 0.1
+      pos_enc_at_attn: false
+      self_attention:
+        _target_: sam2.modeling.sam.transformer.RoPEAttention
+        rope_theta: 10000.0
+        feat_sizes: [64, 64]
+        embedding_dim: 256
+        num_heads: 1
+        downsample_rate: 1
+        dropout: 0.1
+      d_model: 256
+      pos_enc_at_cross_attn_keys: true
+      pos_enc_at_cross_attn_queries: false
+      cross_attention:
+        _target_: sam2.modeling.sam.transformer.RoPEAttention
+        rope_theta: 10000.0
+        feat_sizes: [64, 64]
+        rope_k_repeat: True
+        embedding_dim: 256
+        num_heads: 1
+        downsample_rate: 1
+        dropout: 0.1
+        kv_in_dim: 64
+    num_layers: 4
+
+  memory_encoder:
+      _target_: sam2.modeling.memory_encoder.MemoryEncoder
+      out_dim: 64
+      position_encoding:
+        _target_: sam2.modeling.position_encoding.PositionEmbeddingSine
+        num_pos_feats: 64
+        normalize: true
+        scale: null
+        temperature: 10000
+      mask_downsampler:
+        _target_: sam2.modeling.memory_encoder.MaskDownSampler
+        kernel_size: 3
+        stride: 2
+        padding: 1
+      fuser:
+        _target_: sam2.modeling.memory_encoder.Fuser
+        layer:
+          _target_: sam2.modeling.memory_encoder.CXBlock
+          dim: 256
+          kernel_size: 7
+          padding: 3
+          layer_scale_init_value: 1e-6
+          use_dwconv: True  # depth-wise convs
+        num_layers: 2
+
+  num_maskmem: 7
+  image_size: 1024
+  # apply scaled sigmoid on mask logits for memory encoder, and directly feed input mask as output mask
+  sigmoid_scale_for_mem_enc: 20.0
+  sigmoid_bias_for_mem_enc: -10.0
+  use_mask_input_as_output_without_sam: true
+  # Memory
+  directly_add_no_mem_embed: true
+  no_obj_embed_spatial: true
+  # use high-resolution feature map in the SAM mask decoder
+  use_high_res_features_in_sam: true
+  # output 3 masks on the first click on initial conditioning frames
+  multimask_output_in_sam: true
+  # SAM heads
+  iou_prediction_use_sigmoid: True
+  # cross-attend to object pointers from other frames (based on SAM output tokens) in the encoder
+  use_obj_ptrs_in_encoder: true
+  add_tpos_enc_to_obj_ptrs: true
+  proj_tpos_enc_in_obj_ptrs: true
+  use_signed_tpos_enc_to_obj_ptrs: true
+  only_obj_ptrs_in_the_past_for_eval: true
+  # object occlusion prediction
+  pred_obj_scores: true
+  pred_obj_scores_mlp: true
+  fixed_no_obj_ptr: true
+  # multimask tracking settings
+  multimask_output_for_tracking: true
+  use_multimask_token_for_obj_ptr: true
+  multimask_min_pt_num: 0
+  multimask_max_pt_num: 1
+  use_mlp_for_obj_ptr_proj: true
+  # Compilation flag
+  compile_image_encoder: False

sam2/configs/sam2.1/sam2.1_hiera_t.yaml

@@ -0,0 +1,121 @@
+# @package _global_
+
+# Model
+model:
+  _target_: sam2.modeling.sam2_base.SAM2Base
+  image_encoder:
+    _target_: sam2.modeling.backbones.image_encoder.ImageEncoder
+    scalp: 1
+    trunk:
+      _target_: sam2.modeling.backbones.hieradet.Hiera
+      embed_dim: 96
+      num_heads: 1
+      stages: [1, 2, 7, 2]
+      global_att_blocks: [5, 7, 9]
+      window_pos_embed_bkg_spatial_size: [7, 7]
+    neck:
+      _target_: sam2.modeling.backbones.image_encoder.FpnNeck
+      position_encoding:
+        _target_: sam2.modeling.position_encoding.PositionEmbeddingSine
+        num_pos_feats: 256
+        normalize: true
+        scale: null
+        temperature: 10000
+      d_model: 256
+      backbone_channel_list: [768, 384, 192, 96]
+      fpn_top_down_levels: [2, 3]  # output level 0 and 1 directly use the backbone features
+      fpn_interp_model: nearest
+
+  memory_attention:
+    _target_: sam2.modeling.memory_attention.MemoryAttention
+    d_model: 256
+    pos_enc_at_input: true
+    layer:
+      _target_: sam2.modeling.memory_attention.MemoryAttentionLayer
+      activation: relu
+      dim_feedforward: 2048
+      dropout: 0.1
+      pos_enc_at_attn: false
+      self_attention:
+        _target_: sam2.modeling.sam.transformer.RoPEAttention
+        rope_theta: 10000.0
+        feat_sizes: [64, 64]
+        embedding_dim: 256
+        num_heads: 1
+        downsample_rate: 1
+        dropout: 0.1
+      d_model: 256
+      pos_enc_at_cross_attn_keys: true
+      pos_enc_at_cross_attn_queries: false
+      cross_attention:
+        _target_: sam2.modeling.sam.transformer.RoPEAttention
+        rope_theta: 10000.0
+        feat_sizes: [64, 64]
+        rope_k_repeat: True
+        embedding_dim: 256
+        num_heads: 1
+        downsample_rate: 1
+        dropout: 0.1
+        kv_in_dim: 64
+    num_layers: 4
+
+  memory_encoder:
+      _target_: sam2.modeling.memory_encoder.MemoryEncoder
+      out_dim: 64
+      position_encoding:
+        _target_: sam2.modeling.position_encoding.PositionEmbeddingSine
+        num_pos_feats: 64
+        normalize: true
+        scale: null
+        temperature: 10000
+      mask_downsampler:
+        _target_: sam2.modeling.memory_encoder.MaskDownSampler
+        kernel_size: 3
+        stride: 2
+        padding: 1
+      fuser:
+        _target_: sam2.modeling.memory_encoder.Fuser
+        layer:
+          _target_: sam2.modeling.memory_encoder.CXBlock
+          dim: 256
+          kernel_size: 7
+          padding: 3
+          layer_scale_init_value: 1e-6
+          use_dwconv: True  # depth-wise convs
+        num_layers: 2
+
+  num_maskmem: 7
+  image_size: 1024
+  # apply scaled sigmoid on mask logits for memory encoder, and directly feed input mask as output mask
+  # SAM decoder
+  sigmoid_scale_for_mem_enc: 20.0
+  sigmoid_bias_for_mem_enc: -10.0
+  use_mask_input_as_output_without_sam: true
+  # Memory
+  directly_add_no_mem_embed: true
+  no_obj_embed_spatial: true
+  # use high-resolution feature map in the SAM mask decoder
+  use_high_res_features_in_sam: true
+  # output 3 masks on the first click on initial conditioning frames
+  multimask_output_in_sam: true
+  # SAM heads
+  iou_prediction_use_sigmoid: True
+  # cross-attend to object pointers from other frames (based on SAM output tokens) in the encoder
+  use_obj_ptrs_in_encoder: true
+  add_tpos_enc_to_obj_ptrs: true
+  proj_tpos_enc_in_obj_ptrs: true
+  use_signed_tpos_enc_to_obj_ptrs: true
+  only_obj_ptrs_in_the_past_for_eval: true
+  # object occlusion prediction
+  pred_obj_scores: true
+  pred_obj_scores_mlp: true
+  fixed_no_obj_ptr: true
+  # multimask tracking settings
+  multimask_output_for_tracking: true
+  use_multimask_token_for_obj_ptr: true
+  multimask_min_pt_num: 0
+  multimask_max_pt_num: 1
+  use_mlp_for_obj_ptr_proj: true
+  # Compilation flag
+  # HieraT does not currently support compilation, should always be set to False
+  compile_image_encoder: False

sam2/configs/sam2.1_hiera_b+.yaml

@@ -0,0 +1,137 @@
+# @package _global_
+
+# Model
+model:
+  _target_: sam2.sam2_train.SAM2Train
+  image_encoder:
+    _target_: sam2.modeling.backbones.image_encoder.ImageEncoder
+    scalp: 1
+    trunk:
+      _target_: sam2.modeling.backbones.hieradet.Hiera
+      embed_dim: 112
+      num_heads: 2
+      drop_path_rate: 0.1
+    neck:
+      _target_: sam2.modeling.backbones.image_encoder.FpnNeck
+      position_encoding:
+        _target_: sam2.modeling.position_encoding.PositionEmbeddingSine
+        num_pos_feats: 256
+        normalize: true
+        scale: null
+        temperature: 10000
+      d_model: 256
+      backbone_channel_list: [896, 448, 224, 112]
+      fpn_top_down_levels: [2, 3]  # output level 0 and 1 directly use the backbone features
+      fpn_interp_model: nearest
+
+  memory_attention:
+    _target_: sam2.modeling.memory_attention.MemoryAttention
+    d_model: 256
+    pos_enc_at_input: true
+    layer:
+      _target_: sam2.modeling.memory_attention.MemoryAttentionLayer
+      activation: relu
+      dim_feedforward: 2048
+      dropout: 0.1
+      pos_enc_at_attn: false
+      self_attention:
+        _target_: sam2.modeling.sam.transformer.RoPEAttention
+        rope_theta: 10000.0
+        feat_sizes: [64, 64]
+        embedding_dim: 256
+        num_heads: 1
+        downsample_rate: 1
+        dropout: 0.1
+      d_model: 256
+      pos_enc_at_cross_attn_keys: true
+      pos_enc_at_cross_attn_queries: false
+      cross_attention:
+        _target_: sam2.modeling.sam.transformer.RoPEAttention
+        rope_theta: 10000.0
+        feat_sizes: [64, 64]
+        rope_k_repeat: True
+        embedding_dim: 256
+        num_heads: 1
+        downsample_rate: 1
+        dropout: 0.1
+        kv_in_dim: 64
+    num_layers: 4
+
+  memory_encoder:
+      _target_: sam2.modeling.memory_encoder.MemoryEncoder
+      out_dim: 64
+      position_encoding:
+        _target_: sam2.modeling.position_encoding.PositionEmbeddingSine
+        num_pos_feats: 64
+        normalize: true
+        scale: null
+        temperature: 10000
+      mask_downsampler:
+        _target_: sam2.modeling.memory_encoder.MaskDownSampler
+        kernel_size: 3
+        stride: 2
+        padding: 1
+      fuser:
+        _target_: sam2.modeling.memory_encoder.Fuser
+        layer:
+          _target_: sam2.modeling.memory_encoder.CXBlock
+          dim: 256
+          kernel_size: 7
+          padding: 3
+          layer_scale_init_value: 1e-6
+          use_dwconv: True  # depth-wise convs
+        num_layers: 2
+
+  num_maskmem: 7
+  image_size: 1024
+  # apply scaled sigmoid on mask logits for memory encoder, and directly feed input mask as output mask
+  sigmoid_scale_for_mem_enc: 20.0
+  sigmoid_bias_for_mem_enc: -10.0
+  use_mask_input_as_output_without_sam: true
+  # Memory
+  directly_add_no_mem_embed: true
+  no_obj_embed_spatial: true
+  # use high-resolution feature map in the SAM mask decoder
+  use_high_res_features_in_sam: true
+  # output 3 masks on the first click on initial conditioning frames
+  multimask_output_in_sam: true
+  # SAM heads
+  iou_prediction_use_sigmoid: True
+  # cross-attend to object pointers from other frames (based on SAM output tokens) in the encoder
+  use_obj_ptrs_in_encoder: true
+  add_tpos_enc_to_obj_ptrs: true
+  proj_tpos_enc_in_obj_ptrs: true
+  use_signed_tpos_enc_to_obj_ptrs: true
+  only_obj_ptrs_in_the_past_for_eval: true
+  # object occlusion prediction
+  pred_obj_scores: true
+  pred_obj_scores_mlp: true
+  fixed_no_obj_ptr: true
+  # multimask tracking settings
+  multimask_output_for_tracking: true
+  use_multimask_token_for_obj_ptr: true
+  multimask_min_pt_num: 0
+  multimask_max_pt_num: 1
+  use_mlp_for_obj_ptr_proj: true
+  # Compilation flag
+  compile_image_encoder: False
+
+  ####### Training specific params #######
+  # box/point input and corrections
+  prob_to_use_pt_input_for_train: 0.5
+  prob_to_use_pt_input_for_eval: 0.0
+  prob_to_use_box_input_for_train: 0.5  # 0.5*0.5 = 0.25 prob to use box instead of points
+  prob_to_use_box_input_for_eval: 0.0
+  prob_to_sample_from_gt_for_train: 0.1  # with a small prob, sampling correction points from GT mask instead of prediction errors
+  num_frames_to_correct_for_train: 2  # iteratively sample on random 1~2 frames (always include the first frame)
+  num_frames_to_correct_for_eval: 1  # only iteratively sample on first frame
+  rand_frames_to_correct_for_train: true  # random #init-cond-frame ~ 2
+  add_all_frames_to_correct_as_cond: true  # when a frame receives a correction click, it becomes a conditioning frame (even if it's not initially a conditioning frame)
+  # maximum 2 initial conditioning frames
+  num_init_cond_frames_for_train: 2
+  rand_init_cond_frames_for_train: true  # random 1~2
+  num_correction_pt_per_frame: 7
+  use_act_ckpt_iterative_pt_sampling: false
+
+  num_init_cond_frames_for_eval: 1  # only mask on the first frame
+  forward_backbone_per_frame_for_eval: true

sam2/configs/sam2.1_hiera_l.yaml

@@ -0,0 +1,141 @@
+# @package _global_
+
+# Model
+model:
+  _target_: sam2.sam2_train.SAM2Train
+  image_encoder:
+    _target_: sam2.modeling.backbones.image_encoder.ImageEncoder
+    scalp: 1
+    trunk:
+      _target_: sam2.modeling.backbones.hieradet.Hiera
+      embed_dim: 144
+      num_heads: 2
+      stages: [2, 6, 36, 4]
+      global_att_blocks: [23, 33, 43]
+      window_pos_embed_bkg_spatial_size: [7, 7]
+      window_spec: [8, 4, 16, 8]
+      drop_path_rate: 0.1
+    neck:
+      _target_: sam2.modeling.backbones.image_encoder.FpnNeck
+      position_encoding:
+        _target_: sam2.modeling.position_encoding.PositionEmbeddingSine
+        num_pos_feats: 256
+        normalize: true
+        scale: null
+        temperature: 10000
+      d_model: 256
+      backbone_channel_list: [1152, 576, 288, 144]
+      fpn_top_down_levels: [2, 3]  # output level 0 and 1 directly use the backbone features
+      fpn_interp_model: nearest
+
+  memory_attention:
+    _target_: sam2.modeling.memory_attention.MemoryAttention
+    d_model: 256
+    pos_enc_at_input: true
+    layer:
+      _target_: sam2.modeling.memory_attention.MemoryAttentionLayer
+      activation: relu
+      dim_feedforward: 2048
+      dropout: 0.1
+      pos_enc_at_attn: false
+      self_attention:
+        _target_: sam2.modeling.sam.transformer.RoPEAttention
+        rope_theta: 10000.0
+        feat_sizes: [64, 64]
+        embedding_dim: 256
+        num_heads: 1
+        downsample_rate: 1
+        dropout: 0.1
+      d_model: 256
+      pos_enc_at_cross_attn_keys: true
+      pos_enc_at_cross_attn_queries: false
+      cross_attention:
+        _target_: sam2.modeling.sam.transformer.RoPEAttention
+        rope_theta: 10000.0
+        feat_sizes: [64, 64]
+        rope_k_repeat: True
+        embedding_dim: 256
+        num_heads: 1
+        downsample_rate: 1
+        dropout: 0.1
+        kv_in_dim: 64
+    num_layers: 4
+
+  memory_encoder:
+      _target_: sam2.modeling.memory_encoder.MemoryEncoder
+      out_dim: 64
+      position_encoding:
+        _target_: sam2.modeling.position_encoding.PositionEmbeddingSine
+        num_pos_feats: 64
+        normalize: true
+        scale: null
+        temperature: 10000
+      mask_downsampler:
+        _target_: sam2.modeling.memory_encoder.MaskDownSampler
+        kernel_size: 3
+        stride: 2
+        padding: 1
+      fuser:
+        _target_: sam2.modeling.memory_encoder.Fuser
+        layer:
+          _target_: sam2.modeling.memory_encoder.CXBlock
+          dim: 256
+          kernel_size: 7
+          padding: 3
+          layer_scale_init_value: 1e-6
+          use_dwconv: True  # depth-wise convs
+        num_layers: 2
+
+  num_maskmem: 7
+  image_size: 1024
+  # apply scaled sigmoid on mask logits for memory encoder, and directly feed input mask as output mask
+  sigmoid_scale_for_mem_enc: 20.0
+  sigmoid_bias_for_mem_enc: -10.0
+  use_mask_input_as_output_without_sam: true
+  # Memory
+  directly_add_no_mem_embed: true
+  no_obj_embed_spatial: true
+  # use high-resolution feature map in the SAM mask decoder
+  use_high_res_features_in_sam: true
+  # output 3 masks on the first click on initial conditioning frames
+  multimask_output_in_sam: true
+  # SAM heads
+  iou_prediction_use_sigmoid: True
+  # cross-attend to object pointers from other frames (based on SAM output tokens) in the encoder
+  use_obj_ptrs_in_encoder: true
+  add_tpos_enc_to_obj_ptrs: true
+  proj_tpos_enc_in_obj_ptrs: true
+  use_signed_tpos_enc_to_obj_ptrs: true
+  only_obj_ptrs_in_the_past_for_eval: true
+  # object occlusion prediction
+  pred_obj_scores: true
+  pred_obj_scores_mlp: true
+  fixed_no_obj_ptr: true
+  # multimask tracking settings
+  multimask_output_for_tracking: true
+  use_multimask_token_for_obj_ptr: true
+  multimask_min_pt_num: 0
+  multimask_max_pt_num: 1
+  use_mlp_for_obj_ptr_proj: true
+  # Compilation flag
+  compile_image_encoder: False
+
+  ####### Training specific params #######
+  # box/point input and corrections
+  prob_to_use_pt_input_for_train: 0.5
+  prob_to_use_pt_input_for_eval: 0.0
+  prob_to_use_box_input_for_train: 0.5  # 0.5*0.5 = 0.25 prob to use box instead of points
+  prob_to_use_box_input_for_eval: 0.0
+  prob_to_sample_from_gt_for_train: 0.1  # with a small prob, sampling correction points from GT mask instead of prediction errors
+  num_frames_to_correct_for_train: 2  # iteratively sample on random 1~2 frames (always include the first frame)
+  num_frames_to_correct_for_eval: 1  # only iteratively sample on first frame
+  rand_frames_to_correct_for_train: true  # random #init-cond-frame ~ 2
+  add_all_frames_to_correct_as_cond: true  # when a frame receives a correction click, it becomes a conditioning frame (even if it's not initially a conditioning frame)
+  # maximum 2 initial conditioning frames
+  num_init_cond_frames_for_train: 2
+  rand_init_cond_frames_for_train: true  # random 1~2
+  num_correction_pt_per_frame: 7
+  use_act_ckpt_iterative_pt_sampling: false
+
+  num_init_cond_frames_for_eval: 1  # only mask on the first frame
+  forward_backbone_per_frame_for_eval: true

sam2/configs/sam2.1_hiera_s.yaml

@@ -0,0 +1,140 @@
+# @package _global_
+
+# Model
+model:
+  _target_: sam2.sam2_train.SAM2Train
+  image_encoder:
+    _target_: sam2.modeling.backbones.image_encoder.ImageEncoder
+    scalp: 1
+    trunk:
+      _target_: sam2.modeling.backbones.hieradet.Hiera
+      embed_dim: 96
+      num_heads: 1
+      stages: [1, 2, 11, 2]
+      global_att_blocks: [7, 10, 13]
+      window_pos_embed_bkg_spatial_size: [7, 7]
+      drop_path_rate: 0.1
+    neck:
+      _target_: sam2.modeling.backbones.image_encoder.FpnNeck
+      position_encoding:
+        _target_: sam2.modeling.position_encoding.PositionEmbeddingSine
+        num_pos_feats: 256
+        normalize: true
+        scale: null
+        temperature: 10000
+      d_model: 256
+      backbone_channel_list: [768, 384, 192, 96]
+      fpn_top_down_levels: [2, 3]  # output level 0 and 1 directly use the backbone features
+      fpn_interp_model: nearest
+
+  memory_attention:
+    _target_: sam2.modeling.memory_attention.MemoryAttention
+    d_model: 256
+    pos_enc_at_input: true
+    layer:
+      _target_: sam2.modeling.memory_attention.MemoryAttentionLayer
+      activation: relu
+      dim_feedforward: 2048
+      dropout: 0.1
+      pos_enc_at_attn: false
+      self_attention:
+        _target_: sam2.modeling.sam.transformer.RoPEAttention
+        rope_theta: 10000.0
+        feat_sizes: [64, 64]
+        embedding_dim: 256
+        num_heads: 1
+        downsample_rate: 1
+        dropout: 0.1
+      d_model: 256
+      pos_enc_at_cross_attn_keys: true
+      pos_enc_at_cross_attn_queries: false
+      cross_attention:
+        _target_: sam2.modeling.sam.transformer.RoPEAttention
+        rope_theta: 10000.0
+        feat_sizes: [64, 64]
+        rope_k_repeat: True
+        embedding_dim: 256
+        num_heads: 1
+        downsample_rate: 1
+        dropout: 0.1
+        kv_in_dim: 64
+    num_layers: 4
+
+  memory_encoder:
+      _target_: sam2.modeling.memory_encoder.MemoryEncoder
+      out_dim: 64
+      position_encoding:
+        _target_: sam2.modeling.position_encoding.PositionEmbeddingSine
+        num_pos_feats: 64
+        normalize: true
+        scale: null
+        temperature: 10000
+      mask_downsampler:
+        _target_: sam2.modeling.memory_encoder.MaskDownSampler
+        kernel_size: 3
+        stride: 2
+        padding: 1
+      fuser:
+        _target_: sam2.modeling.memory_encoder.Fuser
+        layer:
+          _target_: sam2.modeling.memory_encoder.CXBlock
+          dim: 256
+          kernel_size: 7
+          padding: 3
+          layer_scale_init_value: 1e-6
+          use_dwconv: True  # depth-wise convs
+        num_layers: 2
+
+  num_maskmem: 7
+  image_size: 1024
+  # apply scaled sigmoid on mask logits for memory encoder, and directly feed input mask as output mask
+  sigmoid_scale_for_mem_enc: 20.0
+  sigmoid_bias_for_mem_enc: -10.0
+  use_mask_input_as_output_without_sam: true
+  # Memory
+  directly_add_no_mem_embed: true
+  no_obj_embed_spatial: true
+  # use high-resolution feature map in the SAM mask decoder
+  use_high_res_features_in_sam: true
+  # output 3 masks on the first click on initial conditioning frames
+  multimask_output_in_sam: true
+  # SAM heads
+  iou_prediction_use_sigmoid: True
+  # cross-attend to object pointers from other frames (based on SAM output tokens) in the encoder
+  use_obj_ptrs_in_encoder: true
+  add_tpos_enc_to_obj_ptrs: true
+  proj_tpos_enc_in_obj_ptrs: true
+  use_signed_tpos_enc_to_obj_ptrs: true
+  only_obj_ptrs_in_the_past_for_eval: true
+  # object occlusion prediction
+  pred_obj_scores: true
+  pred_obj_scores_mlp: true
+  fixed_no_obj_ptr: true
+  # multimask tracking settings
+  multimask_output_for_tracking: true
+  use_multimask_token_for_obj_ptr: true
+  multimask_min_pt_num: 0
+  multimask_max_pt_num: 1
+  use_mlp_for_obj_ptr_proj: true
+  # Compilation flag
+  compile_image_encoder: False
+
+  ####### Training specific params #######
+  # box/point input and corrections
+  prob_to_use_pt_input_for_train: 0.5
+  prob_to_use_pt_input_for_eval: 0.0
+  prob_to_use_box_input_for_train: 0.5  # 0.5*0.5 = 0.25 prob to use box instead of points
+  prob_to_use_box_input_for_eval: 0.0
+  prob_to_sample_from_gt_for_train: 0.1  # with a small prob, sampling correction points from GT mask instead of prediction errors
+  num_frames_to_correct_for_train: 2  # iteratively sample on random 1~2 frames (always include the first frame)
+  num_frames_to_correct_for_eval: 1  # only iteratively sample on first frame
+  rand_frames_to_correct_for_train: true  # random #init-cond-frame ~ 2
+  add_all_frames_to_correct_as_cond: true  # when a frame receives a correction click, it becomes a conditioning frame (even if it's not initially a conditioning frame)
+  # maximum 2 initial conditioning frames
+  num_init_cond_frames_for_train: 2
+  rand_init_cond_frames_for_train: true  # random 1~2
+  num_correction_pt_per_frame: 7
+  use_act_ckpt_iterative_pt_sampling: false
+
+  num_init_cond_frames_for_eval: 1  # only mask on the first frame
+  forward_backbone_per_frame_for_eval: true

sam2/configs/sam2.1_hiera_t.yaml

@@ -0,0 +1,142 @@
+# @package _global_
+
+# Model
+model:
+  _target_: sam2.sam2_train.SAM2Train
+  image_encoder:
+    _target_: sam2.modeling.backbones.image_encoder.ImageEncoder
+    scalp: 1
+    trunk:
+      _target_: sam2.modeling.backbones.hieradet.Hiera
+      embed_dim: 96
+      num_heads: 1
+      stages: [1, 2, 7, 2]
+      global_att_blocks: [5, 7, 9]
+      window_pos_embed_bkg_spatial_size: [7, 7]
+      drop_path_rate: 0.1
+    neck:
+      _target_: sam2.modeling.backbones.image_encoder.FpnNeck
+      position_encoding:
+        _target_: sam2.modeling.position_encoding.PositionEmbeddingSine
+        num_pos_feats: 256
+        normalize: true
+        scale: null
+        temperature: 10000
+      d_model: 256
+      backbone_channel_list: [768, 384, 192, 96]
+      fpn_top_down_levels: [2, 3]  # output level 0 and 1 directly use the backbone features
+      fpn_interp_model: nearest
+
+  memory_attention:
+    _target_: sam2.modeling.memory_attention.MemoryAttention
+    d_model: 256
+    pos_enc_at_input: true
+    layer:
+      _target_: sam2.modeling.memory_attention.MemoryAttentionLayer
+      activation: relu
+      dim_feedforward: 2048
+      dropout: 0.1
+      pos_enc_at_attn: false
+      self_attention:
+        _target_: sam2.modeling.sam.transformer.RoPEAttention
+        rope_theta: 10000.0
+        feat_sizes: [64, 64]
+        embedding_dim: 256
+        num_heads: 1
+        downsample_rate: 1
+        dropout: 0.1
+      d_model: 256
+      pos_enc_at_cross_attn_keys: true
+      pos_enc_at_cross_attn_queries: false
+      cross_attention:
+        _target_: sam2.modeling.sam.transformer.RoPEAttention
+        rope_theta: 10000.0
+        feat_sizes: [64, 64]
+        rope_k_repeat: true
+        embedding_dim: 256
+        num_heads: 1
+        downsample_rate: 1
+        dropout: 0.1
+        kv_in_dim: 64
+    num_layers: 4
+
+  memory_encoder:
+      _target_: sam2.modeling.memory_encoder.MemoryEncoder
+      out_dim: 64
+      position_encoding:
+        _target_: sam2.modeling.position_encoding.PositionEmbeddingSine
+        num_pos_feats: 64
+        normalize: true
+        scale: null
+        temperature: 10000
+      mask_downsampler:
+        _target_: sam2.modeling.memory_encoder.MaskDownSampler
+        kernel_size: 3
+        stride: 2
+        padding: 1
+      fuser:
+        _target_: sam2.modeling.memory_encoder.Fuser
+        layer:
+          _target_: sam2.modeling.memory_encoder.CXBlock
+          dim: 256
+          kernel_size: 7
+          padding: 3
+          layer_scale_init_value: 1e-6
+          use_dwconv: true  # depth-wise convs
+        num_layers: 2
+
+  num_maskmem: 7
+  image_size: 1024
+  # apply scaled sigmoid on mask logits for memory encoder, and directly feed input mask as output mask
+  # SAM decoder
+  sigmoid_scale_for_mem_enc: 20.0
+  sigmoid_bias_for_mem_enc: -10.0
+  use_mask_input_as_output_without_sam: true
+  # Memory
+  directly_add_no_mem_embed: true
+  no_obj_embed_spatial: true
+  # use high-resolution feature map in the SAM mask decoder
+  use_high_res_features_in_sam: true
+  # output 3 masks on the first click on initial conditioning frames
+  multimask_output_in_sam: true
+  # SAM heads
+  iou_prediction_use_sigmoid: true
+  # cross-attend to object pointers from other frames (based on SAM output tokens) in the encoder
+  use_obj_ptrs_in_encoder: true
+  add_tpos_enc_to_obj_ptrs: true
+  proj_tpos_enc_in_obj_ptrs: true
+  use_signed_tpos_enc_to_obj_ptrs: true
+  only_obj_ptrs_in_the_past_for_eval: true
+  # object occlusion prediction
+  pred_obj_scores: true
+  pred_obj_scores_mlp: true
+  fixed_no_obj_ptr: true
+  # multimask tracking settings
+  multimask_output_for_tracking: true
+  use_multimask_token_for_obj_ptr: true
+  multimask_min_pt_num: 0
+  multimask_max_pt_num: 1
+  use_mlp_for_obj_ptr_proj: true
+  # Compilation flag
+  # HieraT does not currently support compilation, should always be set to false
+  compile_image_encoder: false
+
+  ####### Training specific params #######
+  # box/point input and corrections
+  prob_to_use_pt_input_for_train: 0.5
+  prob_to_use_pt_input_for_eval: 0.0
+  prob_to_use_box_input_for_train: 0.5  # 0.5*0.5 = 0.25 prob to use box instead of points
+  prob_to_use_box_input_for_eval: 0.0
+  prob_to_sample_from_gt_for_train: 0.1  # with a small prob, sampling correction points from GT mask instead of prediction errors
+  num_frames_to_correct_for_train: 2  # iteratively sample on random 1~2 frames (always include the first frame)
+  num_frames_to_correct_for_eval: 1  # only iteratively sample on first frame
+  rand_frames_to_correct_for_train: true  # random #init-cond-frame ~ 2
+  add_all_frames_to_correct_as_cond: true  # when a frame receives a correction click, it becomes a conditioning frame (even if it's not initially a conditioning frame)
+  # maximum 2 initial conditioning frames
+  num_init_cond_frames_for_train: 2
+  rand_init_cond_frames_for_train: true  # random 1~2
+  num_correction_pt_per_frame: 7
+  use_act_ckpt_iterative_pt_sampling: false
+
+  num_init_cond_frames_for_eval: 1  # only mask on the first frame
+  forward_backbone_per_frame_for_eval: true

sam2/configs/sam2.1_training/sam2.1_hiera_b+_MOSE_finetune.yaml

@@ -0,0 +1,339 @@
+# @package _global_
+
+scratch:
+  resolution: 1024
+  train_batch_size: 1
+  num_train_workers: 10
+  num_frames: 8
+  max_num_objects: 3
+  base_lr: 5.0e-6
+  vision_lr: 3.0e-06
+  phases_per_epoch: 1
+  num_epochs: 40
+
+dataset:
+  # PATHS to Dataset
+  img_folder: null # PATH to MOSE JPEGImages folder
+  gt_folder: null  # PATH to MOSE Annotations folder
+  file_list_txt: training/assets/MOSE_sample_train_list.txt # Optional PATH to filelist containing a subset of videos to be used for training
+  multiplier: 2
+
+# Video transforms
+vos:
+  train_transforms:
+    - _target_: training.dataset.transforms.ComposeAPI
+      transforms:
+        - _target_: training.dataset.transforms.RandomHorizontalFlip
+          consistent_transform: True
+        - _target_: training.dataset.transforms.RandomAffine
+          degrees: 25
+          shear: 20
+          image_interpolation: bilinear
+          consistent_transform: True
+        - _target_: training.dataset.transforms.RandomResizeAPI
+          sizes: ${scratch.resolution}
+          square: true
+          consistent_transform: True
+        - _target_: training.dataset.transforms.ColorJitter
+          consistent_transform: True
+          brightness: 0.1
+          contrast: 0.03
+          saturation: 0.03
+          hue: null
+        - _target_: training.dataset.transforms.RandomGrayscale
+          p: 0.05
+          consistent_transform: True
+        - _target_: training.dataset.transforms.ColorJitter
+          consistent_transform: False
+          brightness: 0.1
+          contrast: 0.05
+          saturation: 0.05
+          hue: null
+        - _target_: training.dataset.transforms.ToTensorAPI
+        - _target_: training.dataset.transforms.NormalizeAPI
+          mean: [0.485, 0.456, 0.406]
+          std: [0.229, 0.224, 0.225]
+
+trainer:
+  _target_: training.trainer.Trainer
+  mode: train_only
+  max_epochs: ${times:${scratch.num_epochs},${scratch.phases_per_epoch}}
+  accelerator: cuda
+  seed_value: 123
+
+  model:
+    _target_: training.model.sam2.SAM2Train
+    image_encoder:
+      _target_: sam2.modeling.backbones.image_encoder.ImageEncoder
+      scalp: 1
+      trunk:
+        _target_: sam2.modeling.backbones.hieradet.Hiera
+        embed_dim: 112
+        num_heads: 2
+        drop_path_rate: 0.1
+      neck:
+        _target_: sam2.modeling.backbones.image_encoder.FpnNeck
+        position_encoding:
+          _target_: sam2.modeling.position_encoding.PositionEmbeddingSine
+          num_pos_feats: 256
+          normalize: true
+          scale: null
+          temperature: 10000
+        d_model: 256
+        backbone_channel_list: [896, 448, 224, 112]
+        fpn_top_down_levels: [2, 3]  # output level 0 and 1 directly use the backbone features
+        fpn_interp_model: nearest
+
+    memory_attention:
+      _target_: sam2.modeling.memory_attention.MemoryAttention
+      d_model: 256
+      pos_enc_at_input: true
+      layer:
+        _target_: sam2.modeling.memory_attention.MemoryAttentionLayer
+        activation: relu
+        dim_feedforward: 2048
+        dropout: 0.1
+        pos_enc_at_attn: false
+        self_attention:
+          _target_: sam2.modeling.sam.transformer.RoPEAttention
+          rope_theta: 10000.0
+          feat_sizes: [64, 64]
+          embedding_dim: 256
+          num_heads: 1
+          downsample_rate: 1
+          dropout: 0.1
+        d_model: 256
+        pos_enc_at_cross_attn_keys: true
+        pos_enc_at_cross_attn_queries: false
+        cross_attention:
+          _target_: sam2.modeling.sam.transformer.RoPEAttention
+          rope_theta: 10000.0
+          feat_sizes: [64, 64]
+          rope_k_repeat: True
+          embedding_dim: 256
+          num_heads: 1
+          downsample_rate: 1
+          dropout: 0.1
+          kv_in_dim: 64
+      num_layers: 4
+
+    memory_encoder:
+        _target_: sam2.modeling.memory_encoder.MemoryEncoder
+        out_dim: 64
+        position_encoding:
+          _target_: sam2.modeling.position_encoding.PositionEmbeddingSine
+          num_pos_feats: 64
+          normalize: true
+          scale: null
+          temperature: 10000
+        mask_downsampler:
+          _target_: sam2.modeling.memory_encoder.MaskDownSampler
+          kernel_size: 3
+          stride: 2
+          padding: 1
+        fuser:
+          _target_: sam2.modeling.memory_encoder.Fuser
+          layer:
+            _target_: sam2.modeling.memory_encoder.CXBlock
+            dim: 256
+            kernel_size: 7
+            padding: 3
+            layer_scale_init_value: 1e-6
+            use_dwconv: True  # depth-wise convs
+          num_layers: 2
+
+    num_maskmem: 7
+    image_size: ${scratch.resolution}
+    # apply scaled sigmoid on mask logits for memory encoder, and directly feed input mask as output mask
+    sigmoid_scale_for_mem_enc: 20.0
+    sigmoid_bias_for_mem_enc: -10.0
+    use_mask_input_as_output_without_sam: true
+    # Memory
+    directly_add_no_mem_embed: true
+    no_obj_embed_spatial: true
+    # use high-resolution feature map in the SAM mask decoder
+    use_high_res_features_in_sam: true
+    # output 3 masks on the first click on initial conditioning frames
+    multimask_output_in_sam: true
+    # SAM heads
+    iou_prediction_use_sigmoid: True
+    # cross-attend to object pointers from other frames (based on SAM output tokens) in the encoder
+    use_obj_ptrs_in_encoder: true
+    add_tpos_enc_to_obj_ptrs: true
+    proj_tpos_enc_in_obj_ptrs: true
+    use_signed_tpos_enc_to_obj_ptrs: true
+    only_obj_ptrs_in_the_past_for_eval: true
+    # object occlusion prediction
+    pred_obj_scores: true
+    pred_obj_scores_mlp: true
+    fixed_no_obj_ptr: true
+    # multimask tracking settings
+    multimask_output_for_tracking: true
+    use_multimask_token_for_obj_ptr: true
+    multimask_min_pt_num: 0
+    multimask_max_pt_num: 1
+    use_mlp_for_obj_ptr_proj: true
+    # Compilation flag
+    # compile_image_encoder: False
+
+    ####### Training specific params #######
+    # box/point input and corrections
+    prob_to_use_pt_input_for_train: 0.5
+    prob_to_use_pt_input_for_eval: 0.0
+    prob_to_use_box_input_for_train: 0.5  # 0.5*0.5 = 0.25 prob to use box instead of points
+    prob_to_use_box_input_for_eval: 0.0
+    prob_to_sample_from_gt_for_train: 0.1  # with a small prob, sampling correction points from GT mask instead of prediction errors
+    num_frames_to_correct_for_train: 2  # iteratively sample on random 1~2 frames (always include the first frame)
+    num_frames_to_correct_for_eval: 1  # only iteratively sample on first frame
+    rand_frames_to_correct_for_train: True  # random #init-cond-frame ~ 2
+    add_all_frames_to_correct_as_cond: True  # when a frame receives a correction click, it becomes a conditioning frame (even if it's not initially a conditioning frame)
+    # maximum 2 initial conditioning frames
+    num_init_cond_frames_for_train: 2
+    rand_init_cond_frames_for_train: True  # random 1~2
+    num_correction_pt_per_frame: 7
+    use_act_ckpt_iterative_pt_sampling: false
+    
+
+    
+    num_init_cond_frames_for_eval: 1  # only mask on the first frame
+    forward_backbone_per_frame_for_eval: True
+    
+
+  data:
+    train:
+      _target_: training.dataset.sam2_datasets.TorchTrainMixedDataset
+      phases_per_epoch: ${scratch.phases_per_epoch}
+      batch_sizes:
+        - ${scratch.train_batch_size}
+
+      datasets:
+        - _target_: training.dataset.utils.RepeatFactorWrapper
+          dataset:
+            _target_: training.dataset.utils.ConcatDataset
+            datasets:
+            - _target_: training.dataset.vos_dataset.VOSDataset
+              transforms: ${vos.train_transforms}
+              training: true
+              video_dataset:
+                _target_: training.dataset.vos_raw_dataset.PNGRawDataset
+                img_folder: ${dataset.img_folder}
+                gt_folder: ${dataset.gt_folder}
+                file_list_txt: ${dataset.file_list_txt}
+              sampler:
+                _target_: training.dataset.vos_sampler.RandomUniformSampler
+                num_frames: ${scratch.num_frames}
+                max_num_objects: ${scratch.max_num_objects}
+              multiplier: ${dataset.multiplier}
+      shuffle: True
+      num_workers: ${scratch.num_train_workers}
+      pin_memory: True
+      drop_last: True
+      collate_fn:
+        _target_: training.utils.data_utils.collate_fn
+        _partial_: true
+        dict_key: all
+
+  optim:
+    amp:
+      enabled: True
+      amp_dtype: bfloat16
+
+    optimizer:
+      _target_: torch.optim.AdamW
+
+    gradient_clip:
+      _target_: training.optimizer.GradientClipper
+      max_norm: 0.1
+      norm_type: 2
+
+    param_group_modifiers:
+      - _target_: training.optimizer.layer_decay_param_modifier
+        _partial_: True
+        layer_decay_value: 0.9
+        apply_to: 'image_encoder.trunk'
+        overrides:
+          - pattern: '*pos_embed*'
+            value: 1.0
+
+    options:
+      lr:
+        - scheduler:
+            _target_: fvcore.common.param_scheduler.CosineParamScheduler
+            start_value: ${scratch.base_lr}
+            end_value: ${divide:${scratch.base_lr},10}
+        - scheduler:
+            _target_: fvcore.common.param_scheduler.CosineParamScheduler
+            start_value: ${scratch.vision_lr}
+            end_value: ${divide:${scratch.vision_lr},10}
+          param_names:
+            - 'image_encoder.*'
+      weight_decay:
+        - scheduler:
+            _target_: fvcore.common.param_scheduler.ConstantParamScheduler
+            value: 0.1
+        - scheduler:
+            _target_: fvcore.common.param_scheduler.ConstantParamScheduler
+            value: 0.0
+          param_names:
+            - '*bias*'
+          module_cls_names: ['torch.nn.LayerNorm']
+
+  loss:
+    all:
+      _target_: training.loss_fns.MultiStepMultiMasksAndIous
+      weight_dict:
+        loss_mask: 20
+        loss_dice: 1
+        loss_iou: 1
+        loss_class: 1
+      supervise_all_iou: true
+      iou_use_l1_loss: true
+      pred_obj_scores: true
+      focal_gamma_obj_score: 0.0
+      focal_alpha_obj_score: -1.0
+
+  distributed:
+    backend: nccl
+    find_unused_parameters: True
+
+  logging:
+    tensorboard_writer:
+      _target_: training.utils.logger.make_tensorboard_logger
+      log_dir:  ${launcher.experiment_log_dir}/tensorboard
+      flush_secs: 120
+      should_log: True
+    log_dir: ${launcher.experiment_log_dir}/logs
+    log_freq: 10
+
+  # initialize from a SAM 2 checkpoint
+  checkpoint:
+    save_dir: ${launcher.experiment_log_dir}/checkpoints
+    save_freq: 0 # 0 only last checkpoint is saved.
+    model_weight_initializer:
+      _partial_: True
+      _target_: training.utils.checkpoint_utils.load_state_dict_into_model
+      strict: True
+      ignore_unexpected_keys: null
+      ignore_missing_keys: null
+
+      state_dict:
+        _target_: training.utils.checkpoint_utils.load_checkpoint_and_apply_kernels
+        checkpoint_path: ./checkpoints/sam2.1_hiera_base_plus.pt # PATH to SAM 2.1 checkpoint
+        ckpt_state_dict_keys: ['model']
+
+launcher:
+  num_nodes: 1
+  gpus_per_node: 8
+  experiment_log_dir: null # Path to log directory, defaults to ./sam2_logs/${config_name}
+
+# SLURM args if running on a cluster
+submitit:
+  partition: null
+  account: null
+  qos: null
+  cpus_per_task: 10
+  use_cluster: false
+  timeout_hour: 24
+  name: null
+  port_range: [10000, 65000]
+

sam2/configs/sam2/sam2_hiera_b+.yaml

@@ -0,0 +1,113 @@
+# @package _global_
+
+# Model
+model:
+  _target_: sam2.modeling.sam2_base.SAM2Base
+  image_encoder:
+    _target_: sam2.modeling.backbones.image_encoder.ImageEncoder
+    scalp: 1
+    trunk:
+      _target_: sam2.modeling.backbones.hieradet.Hiera
+      embed_dim: 112
+      num_heads: 2
+    neck:
+      _target_: sam2.modeling.backbones.image_encoder.FpnNeck
+      position_encoding:
+        _target_: sam2.modeling.position_encoding.PositionEmbeddingSine
+        num_pos_feats: 256
+        normalize: true
+        scale: null
+        temperature: 10000
+      d_model: 256
+      backbone_channel_list: [896, 448, 224, 112]
+      fpn_top_down_levels: [2, 3]  # output level 0 and 1 directly use the backbone features
+      fpn_interp_model: nearest
+
+  memory_attention:
+    _target_: sam2.modeling.memory_attention.MemoryAttention
+    d_model: 256
+    pos_enc_at_input: true
+    layer:
+      _target_: sam2.modeling.memory_attention.MemoryAttentionLayer
+      activation: relu
+      dim_feedforward: 2048
+      dropout: 0.1
+      pos_enc_at_attn: false
+      self_attention:
+        _target_: sam2.modeling.sam.transformer.RoPEAttention
+        rope_theta: 10000.0
+        feat_sizes: [64, 64]
+        embedding_dim: 256
+        num_heads: 1
+        downsample_rate: 1
+        dropout: 0.1
+      d_model: 256
+      pos_enc_at_cross_attn_keys: true
+      pos_enc_at_cross_attn_queries: false
+      cross_attention:
+        _target_: sam2.modeling.sam.transformer.RoPEAttention
+        rope_theta: 10000.0
+        feat_sizes: [64, 64]
+        rope_k_repeat: True
+        embedding_dim: 256
+        num_heads: 1
+        downsample_rate: 1
+        dropout: 0.1
+        kv_in_dim: 64
+    num_layers: 4
+
+  memory_encoder:
+      _target_: sam2.modeling.memory_encoder.MemoryEncoder
+      out_dim: 64
+      position_encoding:
+        _target_: sam2.modeling.position_encoding.PositionEmbeddingSine
+        num_pos_feats: 64
+        normalize: true
+        scale: null
+        temperature: 10000
+      mask_downsampler:
+        _target_: sam2.modeling.memory_encoder.MaskDownSampler
+        kernel_size: 3
+        stride: 2
+        padding: 1
+      fuser:
+        _target_: sam2.modeling.memory_encoder.Fuser
+        layer:
+          _target_: sam2.modeling.memory_encoder.CXBlock
+          dim: 256
+          kernel_size: 7
+          padding: 3
+          layer_scale_init_value: 1e-6
+          use_dwconv: True  # depth-wise convs
+        num_layers: 2
+
+  num_maskmem: 7
+  image_size: 1024
+  # apply scaled sigmoid on mask logits for memory encoder, and directly feed input mask as output mask
+  sigmoid_scale_for_mem_enc: 20.0
+  sigmoid_bias_for_mem_enc: -10.0
+  use_mask_input_as_output_without_sam: true
+  # Memory
+  directly_add_no_mem_embed: true
+  # use high-resolution feature map in the SAM mask decoder
+  use_high_res_features_in_sam: true
+  # output 3 masks on the first click on initial conditioning frames
+  multimask_output_in_sam: true
+  # SAM heads
+  iou_prediction_use_sigmoid: True
+  # cross-attend to object pointers from other frames (based on SAM output tokens) in the encoder
+  use_obj_ptrs_in_encoder: true
+  add_tpos_enc_to_obj_ptrs: false
+  only_obj_ptrs_in_the_past_for_eval: true
+  # object occlusion prediction
+  pred_obj_scores: true
+  pred_obj_scores_mlp: true
+  fixed_no_obj_ptr: true
+  # multimask tracking settings
+  multimask_output_for_tracking: true
+  use_multimask_token_for_obj_ptr: true
+  multimask_min_pt_num: 0
+  multimask_max_pt_num: 1
+  use_mlp_for_obj_ptr_proj: true
+  # Compilation flag
+  compile_image_encoder: False

sam2/configs/sam2/sam2_hiera_l.yaml

@@ -0,0 +1,117 @@
+# @package _global_
+
+# Model
+model:
+  _target_: sam2.modeling.sam2_base.SAM2Base
+  image_encoder:
+    _target_: sam2.modeling.backbones.image_encoder.ImageEncoder
+    scalp: 1
+    trunk:
+      _target_: sam2.modeling.backbones.hieradet.Hiera
+      embed_dim: 144
+      num_heads: 2
+      stages: [2, 6, 36, 4]
+      global_att_blocks: [23, 33, 43]
+      window_pos_embed_bkg_spatial_size: [7, 7]
+      window_spec: [8, 4, 16, 8]
+    neck:
+      _target_: sam2.modeling.backbones.image_encoder.FpnNeck
+      position_encoding:
+        _target_: sam2.modeling.position_encoding.PositionEmbeddingSine
+        num_pos_feats: 256
+        normalize: true
+        scale: null
+        temperature: 10000
+      d_model: 256
+      backbone_channel_list: [1152, 576, 288, 144]
+      fpn_top_down_levels: [2, 3]  # output level 0 and 1 directly use the backbone features
+      fpn_interp_model: nearest
+
+  memory_attention:
+    _target_: sam2.modeling.memory_attention.MemoryAttention
+    d_model: 256
+    pos_enc_at_input: true
+    layer:
+      _target_: sam2.modeling.memory_attention.MemoryAttentionLayer
+      activation: relu
+      dim_feedforward: 2048
+      dropout: 0.1
+      pos_enc_at_attn: false
+      self_attention:
+        _target_: sam2.modeling.sam.transformer.RoPEAttention
+        rope_theta: 10000.0
+        feat_sizes: [64, 64]
+        embedding_dim: 256
+        num_heads: 1
+        downsample_rate: 1
+        dropout: 0.1
+      d_model: 256
+      pos_enc_at_cross_attn_keys: true
+      pos_enc_at_cross_attn_queries: false
+      cross_attention:
+        _target_: sam2.modeling.sam.transformer.RoPEAttention
+        rope_theta: 10000.0
+        feat_sizes: [64, 64]
+        rope_k_repeat: True
+        embedding_dim: 256
+        num_heads: 1
+        downsample_rate: 1
+        dropout: 0.1
+        kv_in_dim: 64
+    num_layers: 4
+
+  memory_encoder:
+      _target_: sam2.modeling.memory_encoder.MemoryEncoder
+      out_dim: 64
+      position_encoding:
+        _target_: sam2.modeling.position_encoding.PositionEmbeddingSine
+        num_pos_feats: 64
+        normalize: true
+        scale: null
+        temperature: 10000
+      mask_downsampler:
+        _target_: sam2.modeling.memory_encoder.MaskDownSampler
+        kernel_size: 3
+        stride: 2
+        padding: 1
+      fuser:
+        _target_: sam2.modeling.memory_encoder.Fuser
+        layer:
+          _target_: sam2.modeling.memory_encoder.CXBlock
+          dim: 256
+          kernel_size: 7
+          padding: 3
+          layer_scale_init_value: 1e-6
+          use_dwconv: True  # depth-wise convs
+        num_layers: 2
+
+  num_maskmem: 7
+  image_size: 1024
+  # apply scaled sigmoid on mask logits for memory encoder, and directly feed input mask as output mask
+  sigmoid_scale_for_mem_enc: 20.0
+  sigmoid_bias_for_mem_enc: -10.0
+  use_mask_input_as_output_without_sam: true
+  # Memory
+  directly_add_no_mem_embed: true
+  # use high-resolution feature map in the SAM mask decoder
+  use_high_res_features_in_sam: true
+  # output 3 masks on the first click on initial conditioning frames
+  multimask_output_in_sam: true
+  # SAM heads
+  iou_prediction_use_sigmoid: True
+  # cross-attend to object pointers from other frames (based on SAM output tokens) in the encoder
+  use_obj_ptrs_in_encoder: true
+  add_tpos_enc_to_obj_ptrs: false
+  only_obj_ptrs_in_the_past_for_eval: true
+  # object occlusion prediction
+  pred_obj_scores: true
+  pred_obj_scores_mlp: true
+  fixed_no_obj_ptr: true
+  # multimask tracking settings
+  multimask_output_for_tracking: true
+  use_multimask_token_for_obj_ptr: true
+  multimask_min_pt_num: 0
+  multimask_max_pt_num: 1
+  use_mlp_for_obj_ptr_proj: true
+  # Compilation flag
+  compile_image_encoder: False

sam2/configs/sam2/sam2_hiera_s.yaml

@@ -0,0 +1,116 @@
+# @package _global_
+
+# Model
+model:
+  _target_: sam2.modeling.sam2_base.SAM2Base
+  image_encoder:
+    _target_: sam2.modeling.backbones.image_encoder.ImageEncoder
+    scalp: 1
+    trunk:
+      _target_: sam2.modeling.backbones.hieradet.Hiera
+      embed_dim: 96
+      num_heads: 1
+      stages: [1, 2, 11, 2]
+      global_att_blocks: [7, 10, 13]
+      window_pos_embed_bkg_spatial_size: [7, 7]
+    neck:
+      _target_: sam2.modeling.backbones.image_encoder.FpnNeck
+      position_encoding:
+        _target_: sam2.modeling.position_encoding.PositionEmbeddingSine
+        num_pos_feats: 256
+        normalize: true
+        scale: null
+        temperature: 10000
+      d_model: 256
+      backbone_channel_list: [768, 384, 192, 96]
+      fpn_top_down_levels: [2, 3]  # output level 0 and 1 directly use the backbone features
+      fpn_interp_model: nearest
+
+  memory_attention:
+    _target_: sam2.modeling.memory_attention.MemoryAttention
+    d_model: 256
+    pos_enc_at_input: true
+    layer:
+      _target_: sam2.modeling.memory_attention.MemoryAttentionLayer
+      activation: relu
+      dim_feedforward: 2048
+      dropout: 0.1
+      pos_enc_at_attn: false
+      self_attention:
+        _target_: sam2.modeling.sam.transformer.RoPEAttention
+        rope_theta: 10000.0
+        feat_sizes: [64, 64]
+        embedding_dim: 256
+        num_heads: 1
+        downsample_rate: 1
+        dropout: 0.1
+      d_model: 256
+      pos_enc_at_cross_attn_keys: true
+      pos_enc_at_cross_attn_queries: false
+      cross_attention:
+        _target_: sam2.modeling.sam.transformer.RoPEAttention
+        rope_theta: 10000.0
+        feat_sizes: [64, 64]
+        rope_k_repeat: True
+        embedding_dim: 256
+        num_heads: 1
+        downsample_rate: 1
+        dropout: 0.1
+        kv_in_dim: 64
+    num_layers: 4
+
+  memory_encoder:
+      _target_: sam2.modeling.memory_encoder.MemoryEncoder
+      out_dim: 64
+      position_encoding:
+        _target_: sam2.modeling.position_encoding.PositionEmbeddingSine
+        num_pos_feats: 64
+        normalize: true
+        scale: null
+        temperature: 10000
+      mask_downsampler:
+        _target_: sam2.modeling.memory_encoder.MaskDownSampler
+        kernel_size: 3
+        stride: 2
+        padding: 1
+      fuser:
+        _target_: sam2.modeling.memory_encoder.Fuser
+        layer:
+          _target_: sam2.modeling.memory_encoder.CXBlock
+          dim: 256
+          kernel_size: 7
+          padding: 3
+          layer_scale_init_value: 1e-6
+          use_dwconv: True  # depth-wise convs
+        num_layers: 2
+
+  num_maskmem: 7
+  image_size: 1024
+  # apply scaled sigmoid on mask logits for memory encoder, and directly feed input mask as output mask
+  sigmoid_scale_for_mem_enc: 20.0
+  sigmoid_bias_for_mem_enc: -10.0
+  use_mask_input_as_output_without_sam: true
+  # Memory
+  directly_add_no_mem_embed: true
+  # use high-resolution feature map in the SAM mask decoder
+  use_high_res_features_in_sam: true
+  # output 3 masks on the first click on initial conditioning frames
+  multimask_output_in_sam: true
+  # SAM heads
+  iou_prediction_use_sigmoid: True
+  # cross-attend to object pointers from other frames (based on SAM output tokens) in the encoder
+  use_obj_ptrs_in_encoder: true
+  add_tpos_enc_to_obj_ptrs: false
+  only_obj_ptrs_in_the_past_for_eval: true
+  # object occlusion prediction
+  pred_obj_scores: true
+  pred_obj_scores_mlp: true
+  fixed_no_obj_ptr: true
+  # multimask tracking settings
+  multimask_output_for_tracking: true
+  use_multimask_token_for_obj_ptr: true
+  multimask_min_pt_num: 0
+  multimask_max_pt_num: 1
+  use_mlp_for_obj_ptr_proj: true
+  # Compilation flag
+  compile_image_encoder: False

sam2/configs/sam2/sam2_hiera_t.yaml

@@ -0,0 +1,118 @@
+# @package _global_
+
+# Model
+model:
+  _target_: sam2.modeling.sam2_base.SAM2Base
+  image_encoder:
+    _target_: sam2.modeling.backbones.image_encoder.ImageEncoder
+    scalp: 1
+    trunk:
+      _target_: sam2.modeling.backbones.hieradet.Hiera
+      embed_dim: 96
+      num_heads: 1
+      stages: [1, 2, 7, 2]
+      global_att_blocks: [5, 7, 9]
+      window_pos_embed_bkg_spatial_size: [7, 7]
+    neck:
+      _target_: sam2.modeling.backbones.image_encoder.FpnNeck
+      position_encoding:
+        _target_: sam2.modeling.position_encoding.PositionEmbeddingSine
+        num_pos_feats: 256
+        normalize: true
+        scale: null
+        temperature: 10000
+      d_model: 256
+      backbone_channel_list: [768, 384, 192, 96]
+      fpn_top_down_levels: [2, 3]  # output level 0 and 1 directly use the backbone features
+      fpn_interp_model: nearest
+
+  memory_attention:
+    _target_: sam2.modeling.memory_attention.MemoryAttention
+    d_model: 256
+    pos_enc_at_input: true
+    layer:
+      _target_: sam2.modeling.memory_attention.MemoryAttentionLayer
+      activation: relu
+      dim_feedforward: 2048
+      dropout: 0.1
+      pos_enc_at_attn: false
+      self_attention:
+        _target_: sam2.modeling.sam.transformer.RoPEAttention
+        rope_theta: 10000.0
+        feat_sizes: [64, 64]
+        embedding_dim: 256
+        num_heads: 1
+        downsample_rate: 1
+        dropout: 0.1
+      d_model: 256
+      pos_enc_at_cross_attn_keys: true
+      pos_enc_at_cross_attn_queries: false
+      cross_attention:
+        _target_: sam2.modeling.sam.transformer.RoPEAttention
+        rope_theta: 10000.0
+        feat_sizes: [64, 64]
+        rope_k_repeat: True
+        embedding_dim: 256
+        num_heads: 1
+        downsample_rate: 1
+        dropout: 0.1
+        kv_in_dim: 64
+    num_layers: 4
+
+  memory_encoder:
+      _target_: sam2.modeling.memory_encoder.MemoryEncoder
+      out_dim: 64
+      position_encoding:
+        _target_: sam2.modeling.position_encoding.PositionEmbeddingSine
+        num_pos_feats: 64
+        normalize: true
+        scale: null
+        temperature: 10000
+      mask_downsampler:
+        _target_: sam2.modeling.memory_encoder.MaskDownSampler
+        kernel_size: 3
+        stride: 2
+        padding: 1
+      fuser:
+        _target_: sam2.modeling.memory_encoder.Fuser
+        layer:
+          _target_: sam2.modeling.memory_encoder.CXBlock
+          dim: 256
+          kernel_size: 7
+          padding: 3
+          layer_scale_init_value: 1e-6
+          use_dwconv: True  # depth-wise convs
+        num_layers: 2
+
+  num_maskmem: 7
+  image_size: 1024
+  # apply scaled sigmoid on mask logits for memory encoder, and directly feed input mask as output mask
+  # SAM decoder
+  sigmoid_scale_for_mem_enc: 20.0
+  sigmoid_bias_for_mem_enc: -10.0
+  use_mask_input_as_output_without_sam: true
+  # Memory
+  directly_add_no_mem_embed: true
+  # use high-resolution feature map in the SAM mask decoder
+  use_high_res_features_in_sam: true
+  # output 3 masks on the first click on initial conditioning frames
+  multimask_output_in_sam: true
+  # SAM heads
+  iou_prediction_use_sigmoid: True
+  # cross-attend to object pointers from other frames (based on SAM output tokens) in the encoder
+  use_obj_ptrs_in_encoder: true
+  add_tpos_enc_to_obj_ptrs: false
+  only_obj_ptrs_in_the_past_for_eval: true
+  # object occlusion prediction
+  pred_obj_scores: true
+  pred_obj_scores_mlp: true
+  fixed_no_obj_ptr: true
+  # multimask tracking settings
+  multimask_output_for_tracking: true
+  use_multimask_token_for_obj_ptr: true
+  multimask_min_pt_num: 0
+  multimask_max_pt_num: 1
+  use_mlp_for_obj_ptr_proj: true
+  # Compilation flag
+  # HieraT does not currently support compilation, should always be set to False
+  compile_image_encoder: False

sam2/csrc/connected_components.cu

@@ -0,0 +1,289 @@
+// Copyright (c) Meta Platforms, Inc. and affiliates.
+// All rights reserved.
+
+// This source code is licensed under the license found in the
+// LICENSE file in the root directory of this source tree.
+
+// adapted from https://github.com/zsef123/Connected_components_PyTorch
+// with license found in the LICENSE_cctorch file in the root directory.
+#include <ATen/cuda/CUDAContext.h>
+#include <cuda.h>
+#include <cuda_runtime.h>
+#include <torch/extension.h>
+#include <torch/script.h>
+#include <vector>
+
+// 2d
+#define BLOCK_ROWS 16
+#define BLOCK_COLS 16
+
+namespace cc2d {
+
+template <typename T>
+__device__ __forceinline__ unsigned char hasBit(T bitmap, unsigned char pos) {
+  return (bitmap >> pos) & 1;
+}
+
+__device__ int32_t find(const int32_t* s_buf, int32_t n) {
+  while (s_buf[n] != n)
+    n = s_buf[n];
+  return n;
+}
+
+__device__ int32_t find_n_compress(int32_t* s_buf, int32_t n) {
+  const int32_t id = n;
+  while (s_buf[n] != n) {
+    n = s_buf[n];
+    s_buf[id] = n;
+  }
+  return n;
+}
+
+__device__ void union_(int32_t* s_buf, int32_t a, int32_t b) {
+  bool done;
+  do {
+    a = find(s_buf, a);
+    b = find(s_buf, b);
+
+    if (a < b) {
+      int32_t old = atomicMin(s_buf + b, a);
+      done = (old == b);
+      b = old;
+    } else if (b < a) {
+      int32_t old = atomicMin(s_buf + a, b);
+      done = (old == a);
+      a = old;
+    } else
+      done = true;
+
+  } while (!done);
+}
+
+__global__ void
+init_labeling(int32_t* label, const uint32_t W, const uint32_t H) {
+  const uint32_t row = (blockIdx.y * blockDim.y + threadIdx.y) * 2;
+  const uint32_t col = (blockIdx.x * blockDim.x + threadIdx.x) * 2;
+  const uint32_t idx = row * W + col;
+
+  if (row < H && col < W)
+    label[idx] = idx;
+}
+
+__global__ void
+merge(uint8_t* img, int32_t* label, const uint32_t W, const uint32_t H) {
+  const uint32_t row = (blockIdx.y * blockDim.y + threadIdx.y) * 2;
+  const uint32_t col = (blockIdx.x * blockDim.x + threadIdx.x) * 2;
+  const uint32_t idx = row * W + col;
+
+  if (row >= H || col >= W)
+    return;
+
+  uint32_t P = 0;
+
+  if (img[idx])
+    P |= 0x777;
+  if (row + 1 < H && img[idx + W])
+    P |= 0x777 << 4;
+  if (col + 1 < W && img[idx + 1])
+    P |= 0x777 << 1;
+
+  if (col == 0)
+    P &= 0xEEEE;
+  if (col + 1 >= W)
+    P &= 0x3333;
+  else if (col + 2 >= W)
+    P &= 0x7777;
+
+  if (row == 0)
+    P &= 0xFFF0;
+  if (row + 1 >= H)
+    P &= 0xFF;
+
+  if (P > 0) {
+    // If need check about top-left pixel(if flag the first bit) and hit the
+    // top-left pixel
+    if (hasBit(P, 0) && img[idx - W - 1]) {
+      union_(label, idx, idx - 2 * W - 2); // top left block
+    }
+
+    if ((hasBit(P, 1) && img[idx - W]) || (hasBit(P, 2) && img[idx - W + 1]))
+      union_(label, idx, idx - 2 * W); // top bottom block
+
+    if (hasBit(P, 3) && img[idx + 2 - W])
+      union_(label, idx, idx - 2 * W + 2); // top right block
+
+    if ((hasBit(P, 4) && img[idx - 1]) || (hasBit(P, 8) && img[idx + W - 1]))
+      union_(label, idx, idx - 2); // just left block
+  }
+}
+
+__global__ void compression(int32_t* label, const int32_t W, const int32_t H) {
+  const uint32_t row = (blockIdx.y * blockDim.y + threadIdx.y) * 2;
+  const uint32_t col = (blockIdx.x * blockDim.x + threadIdx.x) * 2;
+  const uint32_t idx = row * W + col;
+
+  if (row < H && col < W)
+    find_n_compress(label, idx);
+}
+
+__global__ void final_labeling(
+    const uint8_t* img,
+    int32_t* label,
+    const int32_t W,
+    const int32_t H) {
+  const uint32_t row = (blockIdx.y * blockDim.y + threadIdx.y) * 2;
+  const uint32_t col = (blockIdx.x * blockDim.x + threadIdx.x) * 2;
+  const uint32_t idx = row * W + col;
+
+  if (row >= H || col >= W)
+    return;
+
+  int32_t y = label[idx] + 1;
+
+  if (img[idx])
+    label[idx] = y;
+  else
+    label[idx] = 0;
+
+  if (col + 1 < W) {
+    if (img[idx + 1])
+      label[idx + 1] = y;
+    else
+      label[idx + 1] = 0;
+
+    if (row + 1 < H) {
+      if (img[idx + W + 1])
+        label[idx + W + 1] = y;
+      else
+        label[idx + W + 1] = 0;
+    }
+  }
+
+  if (row + 1 < H) {
+    if (img[idx + W])
+      label[idx + W] = y;
+    else
+      label[idx + W] = 0;
+  }
+}
+
+__global__ void init_counting(
+    const int32_t* label,
+    int32_t* count_init,
+    const int32_t W,
+    const int32_t H) {
+  const uint32_t row = (blockIdx.y * blockDim.y + threadIdx.y);
+  const uint32_t col = (blockIdx.x * blockDim.x + threadIdx.x);
+  const uint32_t idx = row * W + col;
+
+  if (row >= H || col >= W)
+    return;
+
+  int32_t y = label[idx];
+  if (y > 0) {
+    int32_t count_idx = y - 1;
+    atomicAdd(count_init + count_idx, 1);
+  }
+}
+
+__global__ void final_counting(
+    const int32_t* label,
+    const int32_t* count_init,
+    int32_t* count_final,
+    const int32_t W,
+    const int32_t H) {
+  const uint32_t row = (blockIdx.y * blockDim.y + threadIdx.y);
+  const uint32_t col = (blockIdx.x * blockDim.x + threadIdx.x);
+  const uint32_t idx = row * W + col;
+
+  if (row >= H || col >= W)
+    return;
+
+  int32_t y = label[idx];
+  if (y > 0) {
+    int32_t count_idx = y - 1;
+    count_final[idx] = count_init[count_idx];
+  } else {
+    count_final[idx] = 0;
+  }
+}
+
+} // namespace cc2d
+
+std::vector<torch::Tensor> get_connected_componnets(
+    const torch::Tensor& inputs) {
+  AT_ASSERTM(inputs.is_cuda(), "inputs must be a CUDA tensor");
+  AT_ASSERTM(inputs.ndimension() == 4, "inputs must be [N, 1, H, W] shape");
+  AT_ASSERTM(
+      inputs.scalar_type() == torch::kUInt8, "inputs must be a uint8 type");
+
+  const uint32_t N = inputs.size(0);
+  const uint32_t C = inputs.size(1);
+  const uint32_t H = inputs.size(2);
+  const uint32_t W = inputs.size(3);
+
+  AT_ASSERTM(C == 1, "inputs must be [N, 1, H, W] shape");
+  AT_ASSERTM((H % 2) == 0, "height must be an even number");
+  AT_ASSERTM((W % 2) == 0, "width must be an even number");
+
+  // label must be uint32_t
+  auto label_options =
+      torch::TensorOptions().dtype(torch::kInt32).device(inputs.device());
+  torch::Tensor labels = torch::zeros({N, C, H, W}, label_options);
+  torch::Tensor counts_init = torch::zeros({N, C, H, W}, label_options);
+  torch::Tensor counts_final = torch::zeros({N, C, H, W}, label_options);
+
+  dim3 grid = dim3(
+      ((W + 1) / 2 + BLOCK_COLS - 1) / BLOCK_COLS,
+      ((H + 1) / 2 + BLOCK_ROWS - 1) / BLOCK_ROWS);
+  dim3 block = dim3(BLOCK_COLS, BLOCK_ROWS);
+  dim3 grid_count =
+      dim3((W + BLOCK_COLS) / BLOCK_COLS, (H + BLOCK_ROWS) / BLOCK_ROWS);
+  dim3 block_count = dim3(BLOCK_COLS, BLOCK_ROWS);
+  cudaStream_t stream = at::cuda::getCurrentCUDAStream();
+
+  for (int n = 0; n < N; n++) {
+    uint32_t offset = n * H * W;
+
+    cc2d::init_labeling<<<grid, block, 0, stream>>>(
+        labels.data_ptr<int32_t>() + offset, W, H);
+    cc2d::merge<<<grid, block, 0, stream>>>(
+        inputs.data_ptr<uint8_t>() + offset,
+        labels.data_ptr<int32_t>() + offset,
+        W,
+        H);
+    cc2d::compression<<<grid, block, 0, stream>>>(
+        labels.data_ptr<int32_t>() + offset, W, H);
+    cc2d::final_labeling<<<grid, block, 0, stream>>>(
+        inputs.data_ptr<uint8_t>() + offset,
+        labels.data_ptr<int32_t>() + offset,
+        W,
+        H);
+
+    // get the counting of each pixel
+    cc2d::init_counting<<<grid_count, block_count, 0, stream>>>(
+        labels.data_ptr<int32_t>() + offset,
+        counts_init.data_ptr<int32_t>() + offset,
+        W,
+        H);
+    cc2d::final_counting<<<grid_count, block_count, 0, stream>>>(
+        labels.data_ptr<int32_t>() + offset,
+        counts_init.data_ptr<int32_t>() + offset,
+        counts_final.data_ptr<int32_t>() + offset,
+        W,
+        H);
+  }
+
+  // returned values are [labels, counts]
+  std::vector<torch::Tensor> outputs;
+  outputs.push_back(labels);
+  outputs.push_back(counts_final);
+  return outputs;
+}
+
+PYBIND11_MODULE(TORCH_EXTENSION_NAME, m) {
+  m.def(
+      "get_connected_componnets",
+      &get_connected_componnets,
+      "get_connected_componnets");
+}

sam2/loss_fns.py

@@ -0,0 +1,288 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.
+
+from collections import defaultdict
+from typing import Dict, List
+
+import torch
+import torch.distributed
+import torch.nn as nn
+import torch.nn.functional as F
+from nncore.engine import comm
+
+
+def dice_loss(inputs, targets, num_objects, loss_on_multimask=False):
+    """
+    Compute the DICE loss, similar to generalized IOU for masks
+    Args:
+        inputs: A float tensor of arbitrary shape.
+                The predictions for each example.
+        targets: A float tensor with the same shape as inputs. Stores the binary
+                 classification label for each element in inputs
+                (0 for the negative class and 1 for the positive class).
+        num_objects: Number of objects in the batch
+        loss_on_multimask: True if multimask prediction is enabled
+    Returns:
+        Dice loss tensor
+    """
+    inputs = inputs.sigmoid()
+    if loss_on_multimask:
+        # inputs and targets are [N, M, H, W] where M corresponds to multiple predicted masks
+        assert inputs.dim() == 4 and targets.dim() == 4
+        # flatten spatial dimension while keeping multimask channel dimension
+        inputs = inputs.flatten(2)
+        targets = targets.flatten(2)
+        numerator = 2 * (inputs * targets).sum(-1)
+    else:
+        inputs = inputs.flatten(1)
+        numerator = 2 * (inputs * targets).sum(1)
+    denominator = inputs.sum(-1) + targets.sum(-1)
+    loss = 1 - (numerator + 1) / (denominator + 1)
+    if loss_on_multimask:
+        return loss / num_objects
+    return loss.sum() / num_objects
+
+
+def sigmoid_focal_loss(
+    inputs,
+    targets,
+    num_objects,
+    alpha: float = 0.25,
+    gamma: float = 2,
+    loss_on_multimask=False,
+):
+    """
+    Loss used in RetinaNet for dense detection: https://arxiv.org/abs/1708.02002.
+    Args:
+        inputs: A float tensor of arbitrary shape.
+                The predictions for each example.
+        targets: A float tensor with the same shape as inputs. Stores the binary
+                 classification label for each element in inputs
+                (0 for the negative class and 1 for the positive class).
+        num_objects: Number of objects in the batch
+        alpha: (optional) Weighting factor in range (0,1) to balance
+                positive vs negative examples. Default = -1 (no weighting).
+        gamma: Exponent of the modulating factor (1 - p_t) to
+               balance easy vs hard examples.
+        loss_on_multimask: True if multimask prediction is enabled
+    Returns:
+        focal loss tensor
+    """
+    prob = inputs.sigmoid()
+    ce_loss = F.binary_cross_entropy_with_logits(inputs, targets, reduction="none")
+    p_t = prob * targets + (1 - prob) * (1 - targets)
+    loss = ce_loss * ((1 - p_t)**gamma)
+
+    if alpha >= 0:
+        alpha_t = alpha * targets + (1 - alpha) * (1 - targets)
+        loss = alpha_t * loss
+
+    if loss_on_multimask:
+        # loss is [N, M, H, W] where M corresponds to multiple predicted masks
+        assert loss.dim() == 4
+        return loss.flatten(2).mean(-1) / num_objects  # average over spatial dims
+    return loss.mean(1).sum() / num_objects
+
+
+def iou_loss(inputs, targets, pred_ious, num_objects, loss_on_multimask=False, use_l1_loss=False):
+    """
+    Args:
+        inputs: A float tensor of arbitrary shape.
+                The predictions for each example.
+        targets: A float tensor with the same shape as inputs. Stores the binary
+                 classification label for each element in inputs
+                (0 for the negative class and 1 for the positive class).
+        pred_ious: A float tensor containing the predicted IoUs scores per mask
+        num_objects: Number of objects in the batch
+        loss_on_multimask: True if multimask prediction is enabled
+        use_l1_loss: Whether to use L1 loss is used instead of MSE loss
+    Returns:
+        IoU loss tensor
+    """
+    assert inputs.dim() == 4 and targets.dim() == 4
+    pred_mask = inputs.flatten(2) > 0
+    gt_mask = targets.flatten(2) > 0
+    area_i = torch.sum(pred_mask & gt_mask, dim=-1).float()
+    area_u = torch.sum(pred_mask | gt_mask, dim=-1).float()
+    actual_ious = area_i / torch.clamp(area_u, min=1.0)
+
+    if use_l1_loss:
+        loss = F.l1_loss(pred_ious, actual_ious, reduction="none")
+    else:
+        loss = F.mse_loss(pred_ious, actual_ious, reduction="none")
+    if loss_on_multimask:
+        return loss / num_objects
+    return loss.sum() / num_objects
+
+
+class MultiStepMultiMasksAndIous(nn.Module):
+
+    def __init__(
+        self,
+        weight_dict,
+        focal_alpha=0.25,
+        focal_gamma=2,
+        supervise_all_iou=False,
+        iou_use_l1_loss=False,
+        pred_obj_scores=False,
+        focal_gamma_obj_score=0.0,
+        focal_alpha_obj_score=-1,
+    ):
+        """
+        This class computes the multi-step multi-mask and IoU losses.
+        Args:
+            weight_dict: dict containing weights for focal, dice, iou losses
+            focal_alpha: alpha for sigmoid focal loss
+            focal_gamma: gamma for sigmoid focal loss
+            supervise_all_iou: if True, back-prop iou losses for all predicted masks
+            iou_use_l1_loss: use L1 loss instead of MSE loss for iou
+            pred_obj_scores: if True, compute loss for object scores
+            focal_gamma_obj_score: gamma for sigmoid focal loss on object scores
+            focal_alpha_obj_score: alpha for sigmoid focal loss on object scores
+        """
+
+        super().__init__()
+        self.weight_dict = weight_dict
+        self.focal_alpha = focal_alpha
+        self.focal_gamma = focal_gamma
+        assert "loss_mask" in self.weight_dict
+        assert "loss_dice" in self.weight_dict
+        assert "loss_iou" in self.weight_dict
+        if "loss_class" not in self.weight_dict:
+            self.weight_dict["loss_class"] = 0.0
+
+        self.focal_alpha_obj_score = focal_alpha_obj_score
+        self.focal_gamma_obj_score = focal_gamma_obj_score
+        self.supervise_all_iou = supervise_all_iou
+        self.iou_use_l1_loss = iou_use_l1_loss
+        self.pred_obj_scores = pred_obj_scores
+
+    def forward(self, outs_batch: List[Dict], targets_batch: torch.Tensor):
+        assert len(outs_batch) == len(targets_batch)
+        num_objects = torch.tensor((targets_batch.shape[1]), device=targets_batch.device,
+                                   dtype=torch.float)  # Number of objects is fixed within a batch
+        if comm.is_distributed():
+            torch.distributed.all_reduce(num_objects)
+        num_objects = torch.clamp(num_objects / comm.get_world_size(), min=1).item()
+
+        losses = defaultdict(int)
+        for outs, targets in zip(outs_batch, targets_batch):
+            cur_losses = self._forward(outs, targets, num_objects)
+            for k, v in cur_losses.items():
+                losses[k] += v
+
+        return losses
+
+    def _forward(self, outputs: Dict, targets: torch.Tensor, num_objects):
+        """
+        Compute the losses related to the masks: the focal loss and the dice loss.
+        and also the MAE or MSE loss between predicted IoUs and actual IoUs.
+
+        Here "multistep_pred_multimasks_high_res" is a list of multimasks (tensors
+        of shape [N, M, H, W], where M could be 1 or larger, corresponding to
+        one or multiple predicted masks from a click.
+
+        We back-propagate focal, dice losses only on the prediction channel
+        with the lowest focal+dice loss between predicted mask and ground-truth.
+        If `supervise_all_iou` is True, we backpropagate ious losses for all predicted masks.
+        """
+
+        target_masks = targets.unsqueeze(1).float()
+        assert target_masks.dim() == 4  # [N, 1, H, W]
+        src_masks_list = outputs["multistep_pred_multimasks_high_res"]
+        ious_list = outputs["multistep_pred_ious"]
+        object_score_logits_list = outputs["multistep_object_score_logits"]
+
+        assert len(src_masks_list) == len(ious_list)
+        assert len(object_score_logits_list) == len(ious_list)
+
+        # accumulate the loss over prediction steps
+        losses = {"loss_mask": 0, "loss_dice": 0, "loss_iou": 0, "loss_class": 0}
+        for src_masks, ious, object_score_logits in zip(src_masks_list, ious_list, object_score_logits_list):
+            self._update_losses(losses, src_masks, target_masks, ious, num_objects, object_score_logits)
+        losses["core_loss"] = self.reduce_loss(losses)
+        return losses
+
+    def _update_losses(self, losses, src_masks, target_masks, ious, num_objects, object_score_logits):
+        target_masks = target_masks.expand_as(src_masks)
+        # get focal, dice and iou loss on all output masks in a prediction step
+        loss_multimask = sigmoid_focal_loss(
+            src_masks,
+            target_masks,
+            num_objects,
+            alpha=self.focal_alpha,
+            gamma=self.focal_gamma,
+            loss_on_multimask=True,
+        )
+        loss_multidice = dice_loss(src_masks, target_masks, num_objects, loss_on_multimask=True)
+        if not self.pred_obj_scores:
+            loss_class = torch.tensor(0.0, dtype=loss_multimask.dtype, device=loss_multimask.device)
+            target_obj = torch.ones(
+                loss_multimask.shape[0],
+                1,
+                dtype=loss_multimask.dtype,
+                device=loss_multimask.device,
+            )
+        else:
+            target_obj = torch.any((target_masks[:, 0] > 0).flatten(1), dim=-1)[..., None].float()
+            loss_class = sigmoid_focal_loss(
+                object_score_logits,
+                target_obj,
+                num_objects,
+                alpha=self.focal_alpha_obj_score,
+                gamma=self.focal_gamma_obj_score,
+            )
+
+        loss_multiiou = iou_loss(
+            src_masks,
+            target_masks,
+            ious,
+            num_objects,
+            loss_on_multimask=True,
+            use_l1_loss=self.iou_use_l1_loss,
+        )
+        assert loss_multimask.dim() == 2
+        assert loss_multidice.dim() == 2
+        assert loss_multiiou.dim() == 2
+        if loss_multimask.size(1) > 1:
+            # take the mask indices with the smallest focal + dice loss for back propagation
+            loss_combo = (
+                loss_multimask * self.weight_dict["loss_mask"] + loss_multidice * self.weight_dict["loss_dice"])
+            best_loss_inds = torch.argmin(loss_combo, dim=-1)
+            batch_inds = torch.arange(loss_combo.size(0), device=loss_combo.device)
+            loss_mask = loss_multimask[batch_inds, best_loss_inds].unsqueeze(1)
+            loss_dice = loss_multidice[batch_inds, best_loss_inds].unsqueeze(1)
+            # calculate the iou prediction and slot losses only in the index
+            # with the minimum loss for each mask (to be consistent w/ SAM)
+            if self.supervise_all_iou:
+                loss_iou = loss_multiiou.mean(dim=-1).unsqueeze(1)
+            else:
+                loss_iou = loss_multiiou[batch_inds, best_loss_inds].unsqueeze(1)
+        else:
+            loss_mask = loss_multimask
+            loss_dice = loss_multidice
+            loss_iou = loss_multiiou
+
+        # backprop focal, dice and iou loss only if obj present
+        loss_mask = loss_mask * target_obj
+        loss_dice = loss_dice * target_obj
+        loss_iou = loss_iou * target_obj
+
+        # sum over batch dimension (note that the losses are already divided by num_objects)
+        losses["loss_mask"] += loss_mask.sum()
+        losses["loss_dice"] += loss_dice.sum()
+        losses["loss_iou"] += loss_iou.sum()
+        losses["loss_class"] += loss_class
+
+    def reduce_loss(self, losses):
+        reduced_loss = 0.0
+        for loss_key, weight in self.weight_dict.items():
+            if loss_key not in losses:
+                raise ValueError(f"{type(self)} doesn't compute {loss_key}")
+            if weight != 0:
+                reduced_loss += losses[loss_key] * weight
+
+        return reduced_loss

sam2/modeling/__init__.py

@@ -0,0 +1,5 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.

sam2/modeling/backbones/__init__.py

@@ -0,0 +1,5 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.

sam2/modeling/backbones/hieradet.py

@@ -0,0 +1,312 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.
+
+import logging
+from functools import partial
+from typing import List, Tuple, Union
+
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+from iopath.common.file_io import g_pathmgr
+
+from sam2.modeling.backbones.utils import (
+    PatchEmbed,
+    window_partition,
+    window_unpartition,
+)
+
+from sam2.modeling.sam2_utils import DropPath, MLP
+
+
+def do_pool(x: torch.Tensor, pool: nn.Module, norm: nn.Module = None) -> torch.Tensor:
+    if pool is None:
+        return x
+    # (B, H, W, C) -> (B, C, H, W)
+    x = x.permute(0, 3, 1, 2)
+    x = pool(x.float()).to(x.dtype)
+    # (B, C, H', W') -> (B, H', W', C)
+    x = x.permute(0, 2, 3, 1)
+    if norm:
+        x = norm(x)
+
+    return x
+
+
+class MultiScaleAttention(nn.Module):
+
+    def __init__(
+        self,
+        dim: int,
+        dim_out: int,
+        num_heads: int,
+        q_pool: nn.Module = None,
+    ):
+        super().__init__()
+
+        self.dim = dim
+        self.dim_out = dim_out
+        self.num_heads = num_heads
+        self.q_pool = q_pool
+        self.qkv = nn.Linear(dim, dim_out * 3)
+        self.proj = nn.Linear(dim_out, dim_out)
+
+    def forward(self, x: torch.Tensor) -> torch.Tensor:
+        B, H, W, _ = x.shape
+        # qkv with shape (B, H * W, 3, nHead, C)
+        qkv = self.qkv(x).reshape(B, H * W, 3, self.num_heads, -1)
+        # q, k, v with shape (B, H * W, nheads, C)
+        q, k, v = torch.unbind(qkv, 2)
+
+        # Q pooling (for downsample at stage changes)
+        if self.q_pool:
+            q = do_pool(q.reshape(B, H, W, -1), self.q_pool)
+            H, W = q.shape[1:3]  # downsampled shape
+            q = q.reshape(B, H * W, self.num_heads, -1)
+
+        # Torch's SDPA expects [B, nheads, H*W, C] so we transpose
+        x = F.scaled_dot_product_attention(
+            q.transpose(1, 2),
+            k.transpose(1, 2),
+            v.transpose(1, 2),
+        )
+        # Transpose back
+        x = x.transpose(1, 2)
+        x = x.reshape(B, H, W, -1)
+
+        x = self.proj(x)
+
+        return x
+
+
+class MultiScaleBlock(nn.Module):
+
+    def __init__(
+        self,
+        dim: int,
+        dim_out: int,
+        num_heads: int,
+        mlp_ratio: float = 4.0,
+        drop_path: float = 0.0,
+        norm_layer: Union[nn.Module, str] = "LayerNorm",
+        q_stride: Tuple[int, int] = None,
+        act_layer: nn.Module = nn.GELU,
+        window_size: int = 0,
+    ):
+        super().__init__()
+
+        if isinstance(norm_layer, str):
+            norm_layer = partial(getattr(nn, norm_layer), eps=1e-6)
+
+        self.dim = dim
+        self.dim_out = dim_out
+        self.norm1 = norm_layer(dim)
+
+        self.window_size = window_size
+
+        self.pool, self.q_stride = None, q_stride
+        if self.q_stride:
+            self.pool = nn.MaxPool2d(kernel_size=q_stride, stride=q_stride, ceil_mode=False)
+
+        self.attn = MultiScaleAttention(
+            dim,
+            dim_out,
+            num_heads=num_heads,
+            q_pool=self.pool,
+        )
+        self.drop_path = DropPath(drop_path) if drop_path > 0.0 else nn.Identity()
+
+        self.norm2 = norm_layer(dim_out)
+        self.mlp = MLP(
+            dim_out,
+            int(dim_out * mlp_ratio),
+            dim_out,
+            num_layers=2,
+            activation=act_layer,
+        )
+
+        if dim != dim_out:
+            self.proj = nn.Linear(dim, dim_out)
+
+    def forward(self, x: torch.Tensor) -> torch.Tensor:
+        shortcut = x  # B, H, W, C
+        x = self.norm1(x)
+
+        # Skip connection
+        if self.dim != self.dim_out:
+            shortcut = do_pool(self.proj(x), self.pool)
+
+        # Window partition
+        window_size = self.window_size
+        if window_size > 0:
+            H, W = x.shape[1], x.shape[2]
+            x, pad_hw = window_partition(x, window_size)
+
+        # Window Attention + Q Pooling (if stage change)
+        # Apply chunks to reduce memory
+        CHUNK_SIZE, batch_size = 64, x.size(0)
+        if batch_size > CHUNK_SIZE:
+            chunks = []
+            for i in range(0, batch_size, CHUNK_SIZE):
+                chunks.append(self.attn(x[i:i + CHUNK_SIZE]))
+            x = torch.cat(chunks)
+            assert x.size(0) == batch_size
+        else:
+            x = self.attn(x)
+
+        if self.q_stride:
+            # Shapes have changed due to Q pooling
+            window_size = self.window_size // self.q_stride[0]
+            H, W = shortcut.shape[1:3]
+
+            pad_h = (window_size - H % window_size) % window_size
+            pad_w = (window_size - W % window_size) % window_size
+            pad_hw = (H + pad_h, W + pad_w)
+
+        # Reverse window partition
+        if self.window_size > 0:
+            x = window_unpartition(x, window_size, pad_hw, (H, W))
+
+        x = shortcut + self.drop_path(x)
+        # MLP
+        x = x + self.drop_path(self.mlp(self.norm2(x)))
+        return x
+
+
+class Hiera(nn.Module):
+    """
+    Reference: https://arxiv.org/abs/2306.00989
+    """
+
+    def __init__(
+            self,
+            embed_dim: int = 96,  # initial embed dim
+            num_heads: int = 1,  # initial number of heads
+            drop_path_rate: float = 0.0,  # stochastic depth
+            q_pool: int = 3,  # number of q_pool stages
+            q_stride: Tuple[int, int] = (2, 2),  # downsample stride bet. stages
+            stages: Tuple[int, ...] = (2, 3, 16, 3),  # blocks per stage
+            dim_mul: float = 2.0,  # dim_mul factor at stage shift
+            head_mul: float = 2.0,  # head_mul factor at stage shift
+            window_pos_embed_bkg_spatial_size: Tuple[int, int] = (14, 14),
+            # window size per stage, when not using global att.
+            window_spec: Tuple[int, ...] = (
+                8,
+                4,
+                14,
+                7,
+            ),
+            # global attn in these blocks
+            global_att_blocks: Tuple[int, ...] = (
+                12,
+                16,
+                20,
+            ),
+            weights_path=None,
+            return_interm_layers=True,  # return feats from every stage
+    ):
+        super().__init__()
+
+        assert len(stages) == len(window_spec)
+        self.window_spec = window_spec
+
+        depth = sum(stages)
+        self.q_stride = q_stride
+        self.stage_ends = [sum(stages[:i]) - 1 for i in range(1, len(stages) + 1)]
+        assert 0 <= q_pool <= len(self.stage_ends[:-1])
+        self.q_pool_blocks = [x + 1 for x in self.stage_ends[:-1]][:q_pool]
+        self.return_interm_layers = return_interm_layers
+
+        self.patch_embed = PatchEmbed(embed_dim=embed_dim, )
+        # Which blocks have global att?
+        self.global_att_blocks = global_att_blocks
+
+        # Windowed positional embedding (https://arxiv.org/abs/2311.05613)
+        self.window_pos_embed_bkg_spatial_size = window_pos_embed_bkg_spatial_size
+        self.pos_embed = nn.Parameter(torch.zeros(1, embed_dim, *self.window_pos_embed_bkg_spatial_size))
+        self.pos_embed_window = nn.Parameter(torch.zeros(1, embed_dim, self.window_spec[0], self.window_spec[0]))
+
+        dpr = [x.item() for x in torch.linspace(0, drop_path_rate, depth)]  # stochastic depth decay rule
+
+        cur_stage = 1
+        self.blocks = nn.ModuleList()
+
+        for i in range(depth):
+            dim_out = embed_dim
+            # lags by a block, so first block of
+            # next stage uses an initial window size
+            # of previous stage and final window size of current stage
+            window_size = self.window_spec[cur_stage - 1]
+
+            if self.global_att_blocks is not None:
+                window_size = 0 if i in self.global_att_blocks else window_size
+
+            if i - 1 in self.stage_ends:
+                dim_out = int(embed_dim * dim_mul)
+                num_heads = int(num_heads * head_mul)
+                cur_stage += 1
+
+            block = MultiScaleBlock(
+                dim=embed_dim,
+                dim_out=dim_out,
+                num_heads=num_heads,
+                drop_path=dpr[i],
+                q_stride=self.q_stride if i in self.q_pool_blocks else None,
+                window_size=window_size,
+            )
+
+            embed_dim = dim_out
+            self.blocks.append(block)
+
+        self.channel_list = ([self.blocks[i].dim_out
+                              for i in self.stage_ends[::-1]] if return_interm_layers else [self.blocks[-1].dim_out])
+
+        if weights_path is not None:
+            with g_pathmgr.open(weights_path, "rb") as f:
+                chkpt = torch.load(f, map_location="cpu")
+            logging.info("loading Hiera", self.load_state_dict(chkpt, strict=False))
+
+    def _get_pos_embed(self, hw: Tuple[int, int]) -> torch.Tensor:
+        h, w = hw
+        window_embed = self.pos_embed_window
+        pos_embed = F.interpolate(self.pos_embed.float(), size=(h, w), mode="bicubic").to(self.pos_embed.dtype)
+        pos_embed = pos_embed + window_embed.tile([x // y for x, y in zip(pos_embed.shape, window_embed.shape)])
+        pos_embed = pos_embed.permute(0, 2, 3, 1)
+        return pos_embed
+
+    def forward(self, x: torch.Tensor) -> List[torch.Tensor]:
+        x = self.patch_embed(x)
+        # x: (B, H, W, C)
+
+        # Add pos embed
+        x = x + self._get_pos_embed(x.shape[1:3])
+
+        outputs = []
+        for i, blk in enumerate(self.blocks):
+            x = blk(x)
+            if (i == self.stage_ends[-1]) or (i in self.stage_ends and self.return_interm_layers):
+                feats = x.permute(0, 3, 1, 2)
+                outputs.append(feats)
+
+        return outputs
+
+    def get_layer_id(self, layer_name):
+        # https://github.com/microsoft/unilm/blob/master/beit/optim_factory.py#L33
+        num_layers = self.get_num_layers()
+
+        if layer_name.find("rel_pos") != -1:
+            return num_layers + 1
+        elif layer_name.find("pos_embed") != -1:
+            return 0
+        elif layer_name.find("patch_embed") != -1:
+            return 0
+        elif layer_name.find("blocks") != -1:
+            return int(layer_name.split("blocks")[1].split(".")[1]) + 1
+        else:
+            return num_layers + 1
+
+    def get_num_layers(self) -> int:
+        return len(self.blocks)

sam2/modeling/backbones/image_encoder.py

@@ -0,0 +1,145 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.
+
+from typing import List, Optional
+
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+
+
+class ImageEncoder(nn.Module):
+
+    def __init__(
+        self,
+        trunk: nn.Module,
+        neck: nn.Module,
+        scalp: int = 0,
+    ):
+        super().__init__()
+        self.trunk = trunk
+        self.neck = neck
+        self.scalp = scalp
+        assert (
+            self.trunk.channel_list == self.neck.backbone_channel_list
+        ), f"Channel dims of trunk and neck do not match. Trunk: {self.trunk.channel_list}, neck: {self.neck.backbone_channel_list}"
+
+    def forward(self, sample: torch.Tensor):
+        # Forward through backbone
+        # features, pos = self.neck(self.trunk(sample))
+
+        # NOTE: use chunk to reduce memory ------------------------------
+        features, pos, chunk_size = [], [], 16
+        for base_idx in range(0, sample.size(0), chunk_size):
+            chunk_features, chunk_pos = self.neck(self.trunk(sample[base_idx:base_idx + chunk_size]))
+            features.append(chunk_features)
+            pos.append(chunk_pos)
+        features = [torch.cat([e[i] for e in features]) for i in range(len(features[0]))]
+        pos = [torch.cat([e[i] for e in pos]) for i in range(len(pos[0]))]
+        assert features[0].size(0) == pos[0].size(0) == sample.size(0)
+        # ---------------------------------------------------------------
+
+        if self.scalp > 0:
+            # Discard the lowest resolution features
+            features, pos = features[:-self.scalp], pos[:-self.scalp]
+
+        src = features[-1]
+        output = {
+            "vision_features": src,
+            "vision_pos_enc": pos,
+            "backbone_fpn": features,
+        }
+        return output
+
+
+class FpnNeck(nn.Module):
+    """
+    A modified variant of Feature Pyramid Network (FPN) neck
+    (we remove output conv and also do bicubic interpolation similar to ViT
+    pos embed interpolation)
+    """
+
+    def __init__(
+        self,
+        position_encoding: nn.Module,
+        d_model: int,
+        backbone_channel_list: List[int],
+        kernel_size: int = 1,
+        stride: int = 1,
+        padding: int = 0,
+        fpn_interp_model: str = "bilinear",
+        fuse_type: str = "sum",
+        fpn_top_down_levels: Optional[List[int]] = None,
+    ):
+        """Initialize the neck
+        :param trunk: the backbone
+        :param position_encoding: the positional encoding to use
+        :param d_model: the dimension of the model
+        :param neck_norm: the normalization to use
+        """
+        super().__init__()
+        self.position_encoding = position_encoding
+        self.convs = nn.ModuleList()
+        self.backbone_channel_list = backbone_channel_list
+        self.d_model = d_model
+        for dim in backbone_channel_list:
+            current = nn.Sequential()
+            current.add_module(
+                "conv",
+                nn.Conv2d(
+                    in_channels=dim,
+                    out_channels=d_model,
+                    kernel_size=kernel_size,
+                    stride=stride,
+                    padding=padding,
+                ),
+            )
+
+            self.convs.append(current)
+        self.fpn_interp_model = fpn_interp_model
+        assert fuse_type in ["sum", "avg"]
+        self.fuse_type = fuse_type
+
+        # levels to have top-down features in its outputs
+        # e.g. if fpn_top_down_levels is [2, 3], then only outputs of level 2 and 3
+        # have top-down propagation, while outputs of level 0 and level 1 have only
+        # lateral features from the same backbone level.
+        if fpn_top_down_levels is None:
+            # default is to have top-down features on all levels
+            fpn_top_down_levels = range(len(self.convs))
+        self.fpn_top_down_levels = list(fpn_top_down_levels)
+
+    def forward(self, xs: List[torch.Tensor]):
+
+        out = [None] * len(self.convs)
+        pos = [None] * len(self.convs)
+        assert len(xs) == len(self.convs)
+        # fpn forward pass
+        # see https://github.com/facebookresearch/detectron2/blob/main/detectron2/modeling/backbone/fpn.py
+        prev_features = None
+        # forward in top-down order (from low to high resolution)
+        n = len(self.convs) - 1
+        for i in range(n, -1, -1):
+            x = xs[i]
+            lateral_features = self.convs[n - i](x)
+            if i in self.fpn_top_down_levels and prev_features is not None:
+                top_down_features = F.interpolate(
+                    prev_features.float(),
+                    scale_factor=2.0,
+                    mode=self.fpn_interp_model,
+                    align_corners=(None if self.fpn_interp_model == "nearest" else False),
+                    antialias=False,
+                ).to(prev_features.dtype)
+                prev_features = lateral_features + top_down_features
+                if self.fuse_type == "avg":
+                    prev_features /= 2
+            else:
+                prev_features = lateral_features
+            x_out = prev_features
+            out[i] = x_out
+            pos[i] = self.position_encoding(x_out).to(x_out.dtype)
+
+        return out, pos

sam2/modeling/backbones/utils.py

@@ -0,0 +1,88 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.
+"""Some utilities for backbones, in particular for windowing"""
+
+from typing import Tuple
+
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+
+
+def window_partition(x, window_size):
+    """
+    Partition into non-overlapping windows with padding if needed.
+    Args:
+        x (tensor): input tokens with [B, H, W, C].
+        window_size (int): window size.
+    Returns:
+        windows: windows after partition with [B * num_windows, window_size, window_size, C].
+        (Hp, Wp): padded height and width before partition
+    """
+    B, H, W, C = x.shape
+
+    pad_h = (window_size - H % window_size) % window_size
+    pad_w = (window_size - W % window_size) % window_size
+    if pad_h > 0 or pad_w > 0:
+        x = F.pad(x, (0, 0, 0, pad_w, 0, pad_h))
+    Hp, Wp = H + pad_h, W + pad_w
+
+    x = x.view(B, Hp // window_size, window_size, Wp // window_size, window_size, C)
+    windows = x.permute(0, 1, 3, 2, 4, 5).reshape(-1, window_size, window_size, C)
+    return windows, (Hp, Wp)
+
+
+def window_unpartition(windows, window_size, pad_hw, hw):
+    """
+    Window unpartition into original sequences and removing padding.
+    Args:
+        x (tensor): input tokens with [B * num_windows, window_size, window_size, C].
+        window_size (int): window size.
+        pad_hw (Tuple): padded height and width (Hp, Wp).
+        hw (Tuple): original height and width (H, W) before padding.
+    Returns:
+        x: unpartitioned sequences with [B, H, W, C].
+    """
+    Hp, Wp = pad_hw
+    H, W = hw
+    B = windows.shape[0] // (Hp * Wp // window_size // window_size)
+    x = windows.reshape(B, Hp // window_size, Wp // window_size, window_size, window_size, -1)
+    x = x.permute(0, 1, 3, 2, 4, 5).reshape(B, Hp, Wp, -1)
+
+    if Hp > H or Wp > W:
+        x = x[:, :H, :W, :]
+    return x
+
+
+class PatchEmbed(nn.Module):
+    """
+    Image to Patch Embedding.
+    """
+
+    def __init__(
+            self,
+            kernel_size: Tuple[int, ...] = (7, 7),
+            stride: Tuple[int, ...] = (4, 4),
+            padding: Tuple[int, ...] = (3, 3),
+            in_chans: int = 3,
+            embed_dim: int = 768,
+    ):
+        """
+        Args:
+            kernel_size (Tuple): kernel size of the projection layer.
+            stride (Tuple): stride of the projection layer.
+            padding (Tuple): padding size of the projection layer.
+            in_chans (int): Number of input image channels.
+            embed_dim (int):  embed_dim (int): Patch embedding dimension.
+        """
+        super().__init__()
+        self.proj = nn.Conv2d(in_chans, embed_dim, kernel_size=kernel_size, stride=stride, padding=padding)
+
+    def forward(self, x: torch.Tensor) -> torch.Tensor:
+        x = self.proj(x)
+        # B C H W -> B H W C
+        x = x.permute(0, 2, 3, 1)
+        return x

sam2/modeling/memory_attention.py

@@ -0,0 +1,168 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.
+
+from typing import Optional
+
+import torch
+from torch import nn, Tensor
+
+from sam2.modeling.sam.transformer import RoPEAttention
+
+from sam2.modeling.sam2_utils import get_activation_fn, get_clones
+
+
+class MemoryAttentionLayer(nn.Module):
+
+    def __init__(
+        self,
+        activation: str,
+        cross_attention: nn.Module,
+        d_model: int,
+        dim_feedforward: int,
+        dropout: float,
+        pos_enc_at_attn: bool,
+        pos_enc_at_cross_attn_keys: bool,
+        pos_enc_at_cross_attn_queries: bool,
+        self_attention: nn.Module,
+    ):
+        super().__init__()
+        self.d_model = d_model
+        self.dim_feedforward = dim_feedforward
+        self.dropout_value = dropout
+        self.self_attn = self_attention
+        self.cross_attn_image = cross_attention
+
+        # Implementation of Feedforward model
+        self.linear1 = nn.Linear(d_model, dim_feedforward)
+        self.dropout = nn.Dropout(dropout)
+        self.linear2 = nn.Linear(dim_feedforward, d_model)
+
+        self.norm1 = nn.LayerNorm(d_model)
+        self.norm2 = nn.LayerNorm(d_model)
+        self.norm3 = nn.LayerNorm(d_model)
+        self.dropout1 = nn.Dropout(dropout)
+        self.dropout2 = nn.Dropout(dropout)
+        self.dropout3 = nn.Dropout(dropout)
+
+        self.activation_str = activation
+        self.activation = get_activation_fn(activation)
+
+        # Where to add pos enc
+        self.pos_enc_at_attn = pos_enc_at_attn
+        self.pos_enc_at_cross_attn_queries = pos_enc_at_cross_attn_queries
+        self.pos_enc_at_cross_attn_keys = pos_enc_at_cross_attn_keys
+
+    def _forward_sa(self, tgt, query_pos):
+        # Self-Attention
+        tgt2 = self.norm1(tgt)
+        q = k = tgt2 + query_pos if self.pos_enc_at_attn else tgt2
+        tgt2 = self.self_attn(q, k, v=tgt2)
+        tgt = tgt + self.dropout1(tgt2)
+        return tgt
+
+    def _forward_ca(self, tgt, memory, query_pos, pos, num_k_exclude_rope=0):
+        kwds = {}
+        if num_k_exclude_rope > 0:
+            assert isinstance(self.cross_attn_image, RoPEAttention)
+            kwds = {"num_k_exclude_rope": num_k_exclude_rope}
+
+        # Cross-Attention
+        tgt2 = self.norm2(tgt)
+        tgt2 = self.cross_attn_image(
+            q=tgt2 + query_pos if self.pos_enc_at_cross_attn_queries else tgt2,
+            k=memory + pos if self.pos_enc_at_cross_attn_keys else memory,
+            v=memory,
+            **kwds,
+        )
+        tgt = tgt + self.dropout2(tgt2)
+        return tgt
+
+    def forward(
+        self,
+        tgt,
+        memory,
+        pos: Optional[Tensor] = None,
+        query_pos: Optional[Tensor] = None,
+        num_k_exclude_rope: int = 0,
+    ) -> torch.Tensor:
+
+        # Self-Attn, Cross-Attn
+        tgt = self._forward_sa(tgt, query_pos)
+        tgt = self._forward_ca(tgt, memory, query_pos, pos, num_k_exclude_rope)
+        # MLP
+        tgt2 = self.norm3(tgt)
+        tgt2 = self.linear2(self.dropout(self.activation(self.linear1(tgt2))))
+        tgt = tgt + self.dropout3(tgt2)
+        return tgt
+
+
+class MemoryAttention(nn.Module):
+
+    def __init__(
+            self,
+            d_model: int,
+            pos_enc_at_input: bool,
+            layer: nn.Module,
+            num_layers: int,
+            batch_first: bool = True,  # Do layers expect batch first input?
+    ):
+        super().__init__()
+        self.d_model = d_model
+        self.layers = get_clones(layer, num_layers)
+        self.num_layers = num_layers
+        self.norm = nn.LayerNorm(d_model)
+        self.pos_enc_at_input = pos_enc_at_input
+        self.batch_first = batch_first
+
+    def forward(
+            self,
+            curr: torch.Tensor,  # self-attention inputs
+            memory: torch.Tensor,  # cross-attention inputs
+            curr_pos: Optional[Tensor] = None,  # pos_enc for self-attention inputs
+            memory_pos: Optional[Tensor] = None,  # pos_enc for cross-attention inputs
+            num_obj_ptr_tokens: int = 0,  # number of object pointer *tokens*
+    ):
+        if isinstance(curr, list):
+            assert isinstance(curr_pos, list)
+            assert len(curr) == len(curr_pos) == 1
+            curr, curr_pos = (
+                curr[0],
+                curr_pos[0],
+            )
+
+        assert (curr.shape[1] == memory.shape[1]), "Batch size must be the same for curr and memory"
+
+        output = curr
+        if self.pos_enc_at_input and curr_pos is not None:
+            output = output + 0.1 * curr_pos
+
+        if self.batch_first:
+            # Convert to batch first
+            output = output.transpose(0, 1)
+            curr_pos = curr_pos.transpose(0, 1)
+            memory = memory.transpose(0, 1)
+            memory_pos = memory_pos.transpose(0, 1)
+
+        for layer in self.layers:
+            kwds = {}
+            if isinstance(layer.cross_attn_image, RoPEAttention):
+                kwds = {"num_k_exclude_rope": num_obj_ptr_tokens}
+
+            output = layer(
+                tgt=output,
+                memory=memory,
+                pos=memory_pos,
+                query_pos=curr_pos,
+                **kwds,
+            )
+        normed_output = self.norm(output)
+
+        if self.batch_first:
+            # Convert back to seq first
+            normed_output = normed_output.transpose(0, 1)
+            curr_pos = curr_pos.transpose(0, 1)
+
+        return normed_output

sam2/modeling/memory_encoder.py

@@ -0,0 +1,180 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.
+
+import math
+from typing import Tuple
+
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+
+from sam2.modeling.sam2_utils import DropPath, get_clones, LayerNorm2d
+
+
+class MaskDownSampler(nn.Module):
+    """
+    Progressively downsample a mask by total_stride, each time by stride.
+    Note that LayerNorm is applied per *token*, like in ViT.
+
+    With each downsample (by a factor stride**2), channel capacity increases by the same factor.
+    In the end, we linearly project to embed_dim channels.
+    """
+
+    def __init__(
+        self,
+        embed_dim=256,
+        kernel_size=4,
+        stride=4,
+        padding=0,
+        total_stride=16,
+        activation=nn.GELU,
+    ):
+        super().__init__()
+        num_layers = int(math.log2(total_stride) // math.log2(stride))
+        assert stride**num_layers == total_stride
+        self.encoder = nn.Sequential()
+        mask_in_chans, mask_out_chans = 1, 1
+        for _ in range(num_layers):
+            mask_out_chans = mask_in_chans * (stride**2)
+            self.encoder.append(
+                nn.Conv2d(
+                    mask_in_chans,
+                    mask_out_chans,
+                    kernel_size=kernel_size,
+                    stride=stride,
+                    padding=padding,
+                ))
+            self.encoder.append(LayerNorm2d(mask_out_chans))
+            self.encoder.append(activation())
+            mask_in_chans = mask_out_chans
+
+        self.encoder.append(nn.Conv2d(mask_out_chans, embed_dim, kernel_size=1))
+
+    def forward(self, x):
+        return self.encoder(x)
+
+
+# Lightly adapted from ConvNext (https://github.com/facebookresearch/ConvNeXt)
+class CXBlock(nn.Module):
+    r"""ConvNeXt Block. There are two equivalent implementations:
+    (1) DwConv -> LayerNorm (channels_first) -> 1x1 Conv -> GELU -> 1x1 Conv; all in (N, C, H, W)
+    (2) DwConv -> Permute to (N, H, W, C); LayerNorm (channels_last) -> Linear -> GELU -> Linear; Permute back
+    We use (2) as we find it slightly faster in PyTorch
+
+    Args:
+        dim (int): Number of input channels.
+        drop_path (float): Stochastic depth rate. Default: 0.0
+        layer_scale_init_value (float): Init value for Layer Scale. Default: 1e-6.
+    """
+
+    def __init__(
+        self,
+        dim,
+        kernel_size=7,
+        padding=3,
+        drop_path=0.0,
+        layer_scale_init_value=1e-6,
+        use_dwconv=True,
+    ):
+        super().__init__()
+        self.dwconv = nn.Conv2d(
+            dim,
+            dim,
+            kernel_size=kernel_size,
+            padding=padding,
+            groups=dim if use_dwconv else 1,
+        )  # depthwise conv
+        self.norm = LayerNorm2d(dim, eps=1e-6)
+        self.pwconv1 = nn.Linear(dim, 4 * dim)  # pointwise/1x1 convs, implemented with linear layers
+        self.act = nn.GELU()
+        self.pwconv2 = nn.Linear(4 * dim, dim)
+        # NOTE: changed from gamma to weight
+        # https://github.com/huggingface/transformers/issues/29554
+        self.weight = (
+            nn.Parameter(layer_scale_init_value * torch.ones(
+                (dim)), requires_grad=True) if layer_scale_init_value > 0 else None)
+        self.drop_path = DropPath(drop_path) if drop_path > 0.0 else nn.Identity()
+
+    def forward(self, x):
+        input = x
+        x = self.dwconv(x)
+        x = self.norm(x)
+        x = x.permute(0, 2, 3, 1)  # (N, C, H, W) -> (N, H, W, C)
+        x = self.pwconv1(x)
+        x = self.act(x)
+        x = self.pwconv2(x)
+        if self.weight is not None:
+            x = self.weight * x
+        x = x.permute(0, 3, 1, 2)  # (N, H, W, C) -> (N, C, H, W)
+
+        x = input + self.drop_path(x)
+        return x
+
+
+class Fuser(nn.Module):
+
+    def __init__(self, layer, num_layers, dim=None, input_projection=False):
+        super().__init__()
+        self.proj = nn.Identity()
+        self.layers = get_clones(layer, num_layers)
+
+        if input_projection:
+            assert dim is not None
+            self.proj = nn.Conv2d(dim, dim, kernel_size=1)
+
+    def forward(self, x):
+        # normally x: (N, C, H, W)
+        x = self.proj(x)
+        for layer in self.layers:
+            x = layer(x)
+        return x
+
+
+class MemoryEncoder(nn.Module):
+
+    def __init__(
+            self,
+            out_dim,
+            mask_downsampler,
+            fuser,
+            position_encoding,
+            in_dim=256,  # in_dim of pix_feats
+    ):
+        super().__init__()
+
+        self.mask_downsampler = mask_downsampler
+
+        self.pix_feat_proj = nn.Conv2d(in_dim, in_dim, kernel_size=1)
+        self.fuser = fuser
+        self.position_encoding = position_encoding
+        self.out_proj = nn.Identity()
+        if out_dim != in_dim:
+            self.out_proj = nn.Conv2d(in_dim, out_dim, kernel_size=1)
+
+    def forward(
+        self,
+        pix_feat: torch.Tensor,
+        masks: torch.Tensor,
+        skip_mask_sigmoid: bool = False,
+    ) -> Tuple[torch.Tensor, torch.Tensor]:
+        # Process masks
+        # sigmoid, so that less domain shift from gt masks which are bool
+        if not skip_mask_sigmoid:
+            masks = F.sigmoid(masks)
+        masks = self.mask_downsampler(masks)
+
+        # Fuse pix_feats and downsampled masks
+        # in case the visual features are on CPU, cast them to CUDA
+        pix_feat = pix_feat.to(masks.device)
+
+        x = self.pix_feat_proj(pix_feat)
+        x = x + masks
+        x = self.fuser(x)
+        x = self.out_proj(x)
+
+        pos = self.position_encoding(x).to(x.dtype)
+
+        return {"vision_features": x, "vision_pos_enc": [pos]}

sam2/modeling/position_encoding.py

@@ -0,0 +1,312 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.
+
+import math
+from typing import Optional, Tuple
+
+import numpy as np
+import torch
+from torch import nn
+
+
+class PositionEmbeddingSine(nn.Module):
+    """
+    This is a more standard version of the position embedding, very similar to the one
+    used by the Attention Is All You Need paper, generalized to work on images.
+    """
+
+    def __init__(
+            self,
+            num_pos_feats,
+            temperature: int = 10000,
+            normalize: bool = True,
+            scale: Optional[float] = None,
+            # Following settings only relevant
+            # for warmping up cache for compilation
+            warmup_cache: bool = True,
+            image_size: int = 1024,
+            strides: Tuple[int] = (4, 8, 16, 32),
+    ):
+        super().__init__()
+        assert num_pos_feats % 2 == 0, "Expecting even model width"
+        self.num_pos_feats = num_pos_feats // 2
+        self.temperature = temperature
+        self.normalize = normalize
+        if scale is not None and normalize is False:
+            raise ValueError("normalize should be True if scale is passed")
+        if scale is None:
+            scale = 2 * math.pi
+        self.scale = scale
+
+        self.cache = {}
+        if warmup_cache:
+            # Warmup cache for cuda and npu, to help with compilation
+            try:
+                import torch_npu
+                has_npu = torch_npu.npu.is_available()
+            except ImportError:
+                has_npu = False
+            if torch.cuda.is_available() or has_npu:
+                device = torch.device("cuda" if torch.cuda.is_available() else "npu")
+                for stride in strides:
+                    cache_key = (image_size // stride, image_size // stride)
+                    self._pe(1, device, None, *cache_key)
+
+    def _encode_xy(self, x, y):
+        # NOTE: disable autocasting here
+        raise NotImplementedError
+        # The positions are expected to be normalized
+        assert len(x) == len(y) and x.ndim == y.ndim == 1
+        x_embed = x * self.scale
+        y_embed = y * self.scale
+
+        dim_t = torch.arange(self.num_pos_feats, dtype=torch.float32, device=x.device)
+        dim_t = self.temperature**(2 * (dim_t // 2) / self.num_pos_feats)
+
+        pos_x = x_embed[:, None] / dim_t
+        pos_y = y_embed[:, None] / dim_t
+        pos_x = torch.stack((pos_x[:, 0::2].sin(), pos_x[:, 1::2].cos()), dim=2).flatten(1)
+        pos_y = torch.stack((pos_y[:, 0::2].sin(), pos_y[:, 1::2].cos()), dim=2).flatten(1)
+        return pos_x, pos_y
+
+    @torch.no_grad()
+    def encode_boxes(self, x, y, w, h):
+        # NOTE: disable autocasting here
+        raise NotImplementedError
+        pos_x, pos_y = self._encode_xy(x, y)
+        pos = torch.cat((pos_y, pos_x, h[:, None], w[:, None]), dim=1)
+        return pos
+
+    encode = encode_boxes  # Backwards compatibility
+
+    @torch.no_grad()
+    def encode_points(self, x, y, labels):
+        # NOTE: disable autocasting here
+        raise NotImplementedError
+        (bx, nx), (by, ny), (bl, nl) = x.shape, y.shape, labels.shape
+        assert bx == by and nx == ny and bx == bl and nx == nl
+        pos_x, pos_y = self._encode_xy(x.flatten(), y.flatten())
+        pos_x, pos_y = pos_x.reshape(bx, nx, -1), pos_y.reshape(by, ny, -1)
+        pos = torch.cat((pos_y, pos_x, labels[:, :, None]), dim=2)
+        return pos
+
+    @torch.no_grad()
+    def _pe(self, B, device, dtype, *cache_key):
+        H, W = cache_key
+        if cache_key in self.cache:
+            return self.cache[cache_key].to(device)[None].repeat(B, 1, 1, 1)
+
+        # Force fp32 (https://github.com/huggingface/transformers/pull/29285)
+        with torch.autocast(device_type=device.type, enabled=False):
+            y_embed = torch.arange(1, H + 1, dtype=torch.float32, device=device).view(1, -1, 1).repeat(B, 1, W)
+            x_embed = torch.arange(1, W + 1, dtype=torch.float32, device=device).view(1, 1, -1).repeat(B, H, 1)
+
+            if self.normalize:
+                eps = 1e-6
+                y_embed = y_embed / (y_embed[:, -1:, :] + eps) * self.scale
+                x_embed = x_embed / (x_embed[:, :, -1:] + eps) * self.scale
+
+            dim_t = torch.arange(self.num_pos_feats, dtype=torch.float32, device=device)
+            dim_t = self.temperature**(2 * (dim_t // 2) / self.num_pos_feats)
+
+            pos_x = x_embed[:, :, :, None] / dim_t
+            pos_y = y_embed[:, :, :, None] / dim_t
+            pos_x = torch.stack((pos_x[:, :, :, 0::2].sin(), pos_x[:, :, :, 1::2].cos()), dim=4).flatten(3)
+            pos_y = torch.stack((pos_y[:, :, :, 0::2].sin(), pos_y[:, :, :, 1::2].cos()), dim=4).flatten(3)
+            pos = torch.cat((pos_y, pos_x), dim=3).permute(0, 3, 1, 2)
+
+        if dtype is not None:
+            pos = pos.to(dtype)
+
+        self.cache[cache_key] = pos[0]
+        return pos
+
+    @torch.no_grad()
+    def forward(self, x: torch.Tensor):
+        B = x.shape[0]
+        cache_key = (x.shape[-2], x.shape[-1])
+        return self._pe(B, x.device, x.dtype, *cache_key)
+
+
+class PositionEmbeddingRandom(nn.Module):
+    """
+    Positional encoding using random spatial frequencies.
+    """
+
+    def __init__(self, num_pos_feats: int = 64, scale: Optional[float] = None) -> None:
+        super().__init__()
+        if scale is None or scale <= 0.0:
+            scale = 1.0
+        self.register_buffer(
+            "positional_encoding_gaussian_matrix",
+            scale * torch.randn((2, num_pos_feats)),
+        )
+
+    @torch.no_grad()
+    def _pe_encoding(self, coords: torch.Tensor) -> torch.Tensor:
+        """Positionally encode points that are normalized to [0,1]."""
+        # assuming coords are in [0, 1]^2 square and have d_1 x ... x d_n x 2 shape
+        coords = 2 * coords - 1
+        coords = coords @ self.positional_encoding_gaussian_matrix.to(coords.dtype)
+        coords = 2 * np.pi * coords
+        # outputs d_1 x ... x d_n x C shape
+        return torch.cat([torch.sin(coords), torch.cos(coords)], dim=-1)
+
+    @torch.no_grad()
+    def forward(self, size: Tuple[int, int]) -> torch.Tensor:
+        """Generate positional encoding for a grid of the specified size."""
+        h, w = size
+        device = self.positional_encoding_gaussian_matrix.device
+
+        # Force fp32 (https://github.com/huggingface/transformers/pull/29285)
+        with torch.autocast(device_type=device.type, enabled=False):
+            grid = torch.ones((h, w), device=device, dtype=torch.float32)
+            y_embed = grid.cumsum(dim=0) - 0.5
+            x_embed = grid.cumsum(dim=1) - 0.5
+            y_embed = y_embed / h
+            x_embed = x_embed / w
+            pe = self._pe_encoding(torch.stack([x_embed, y_embed], dim=-1))
+
+        pe = pe.to(self.positional_encoding_gaussian_matrix.dtype)
+        return pe.permute(2, 0, 1)  # C x H x W
+
+    @torch.no_grad()
+    def forward_with_coords(self, coords_input: torch.Tensor, image_size: Tuple[int, int]) -> torch.Tensor:
+        """Positionally encode points that are not normalized to [0,1]."""
+        assert coords_input.dtype == torch.float, 'coords_input must be in float32'
+
+        # Force fp32 (https://github.com/huggingface/transformers/pull/29285)
+        with torch.autocast(device_type=coords_input.device.type, enabled=False):
+            coords = coords_input.clone()
+            coords[:, :, 0] = coords[:, :, 0] / image_size[1]
+            coords[:, :, 1] = coords[:, :, 1] / image_size[0]
+            pe = self._pe_encoding(coords.to(torch.float))  # B x N x C
+
+        pe = pe.to(self.positional_encoding_gaussian_matrix.dtype)
+        return pe
+
+
+class PositionEmbedding1DRandom(nn.Module):
+    """
+    Positional encoding using random frequencies for 1D inputs.
+    """
+
+    def __init__(self, num_pos_feats: int = 64, scale: Optional[float] = None) -> None:
+        super().__init__()
+        if scale is None or scale <= 0.0:
+            scale = 1.0
+        self.register_buffer(
+            "positional_encoding_gaussian_matrix",
+            scale * torch.randn((1, num_pos_feats)),
+        )
+
+    @torch.no_grad()
+    def _pe_encoding(self, coords: torch.Tensor) -> torch.Tensor:
+        """Positionally encode points that are normalized to [0,1]."""
+        coords = 2 * coords - 1
+        coords = coords @ self.positional_encoding_gaussian_matrix.to(coords.dtype)
+        coords = 2 * np.pi * coords
+        return torch.cat([torch.sin(coords), torch.cos(coords)], dim=-1)
+
+    @torch.no_grad()
+    def forward(self, size: int) -> torch.Tensor:
+        """Generate positional encoding for a sequence of the specified length."""
+        device = self.positional_encoding_gaussian_matrix.device
+
+        # Force fp32 (https://github.com/huggingface/transformers/pull/29285)
+        with torch.autocast(device_type=device.type, enabled=False):
+            positions = torch.arange(size, device=device, dtype=torch.float32)
+            positions = positions / (size - 1)
+            positions = positions.unsqueeze(-1)
+            pe = self._pe_encoding(positions)
+
+        pe = pe.to(self.positional_encoding_gaussian_matrix.dtype)
+        return pe.permute(1, 0)  # C x L
+
+    @torch.no_grad()
+    def forward_with_coords(self, coords_input: torch.Tensor, seq_length: int) -> torch.Tensor:
+        """Positionally encode raw coordinates by normalizing to [0,1]."""
+        assert coords_input.dtype == torch.float, 'coords_input must be in float32'
+
+        # Force fp32 (https://github.com/huggingface/transformers/pull/29285)
+        with torch.autocast(device_type=coords_input.device.type, enabled=False):
+            coords = coords_input.clone()
+            coords = coords / (seq_length - 1)
+            if coords.dim() == 2:
+                coords = coords.unsqueeze(-1)
+            pe = self._pe_encoding(coords.to(torch.float))  # B x N x C
+
+        pe = pe.to(self.positional_encoding_gaussian_matrix.dtype)
+        return pe
+
+
+# Rotary Positional Encoding, adapted from:
+# 1. https://github.com/meta-llama/codellama/blob/main/llama/model.py
+# 2. https://github.com/naver-ai/rope-vit
+# 3. https://github.com/lucidrains/rotary-embedding-torch
+
+
+@torch.no_grad()
+def init_t_xy(end_x: int, end_y: int):
+    t = torch.arange(end_x * end_y, dtype=torch.float32)
+    t_x = (t % end_x).float()
+    t_y = torch.div(t, end_x, rounding_mode="floor").float()
+    return t_x, t_y
+
+
+@torch.no_grad()
+def compute_axial_cis(dim: int, end_x: int, end_y: int, theta: float = 10000.0):
+    # Force fp32 on CPU (see https://github.com/huggingface/transformers/pull/29285)
+    with torch.autocast(device_type='cpu', enabled=False):
+        freqs_x = 1.0 / (theta**(torch.arange(0, dim, 4)[:(dim // 4)].float() / dim))
+        freqs_y = 1.0 / (theta**(torch.arange(0, dim, 4)[:(dim // 4)].float() / dim))
+
+        t_x, t_y = init_t_xy(end_x, end_y)
+        freqs_x = torch.outer(t_x, freqs_x)
+        freqs_y = torch.outer(t_y, freqs_y)
+        freqs_cis_x = torch.polar(torch.ones_like(freqs_x), freqs_x)
+        freqs_cis_y = torch.polar(torch.ones_like(freqs_y), freqs_y)
+
+    return torch.cat([freqs_cis_x, freqs_cis_y], dim=-1)
+
+
+@torch.no_grad()
+def reshape_for_broadcast(freqs_cis: torch.Tensor, x: torch.Tensor):
+    ndim = x.ndim
+    assert 0 <= 1 < ndim
+    assert freqs_cis.shape == (x.shape[-2], x.shape[-1])
+    shape = [d if i >= ndim - 2 else 1 for i, d in enumerate(x.shape)]
+    return freqs_cis.view(*shape)
+
+
+@torch.no_grad()
+def apply_rotary_enc(
+    xq: torch.Tensor,
+    xk: torch.Tensor,
+    freqs_cis: torch.Tensor,
+    repeat_freqs_k: bool = False,
+):
+    # Force fp32 (https://github.com/huggingface/transformers/pull/29285)
+    with torch.autocast(device_type=freqs_cis.device.type, enabled=False):
+        xq_ = torch.view_as_complex(xq.float().reshape(*xq.shape[:-1], -1, 2))
+        xk_ = (torch.view_as_complex(xk.float().reshape(*xk.shape[:-1], -1, 2)) if xk.shape[-2] != 0 else None)
+        freqs_cis = reshape_for_broadcast(freqs_cis, xq_)
+        xq_out = torch.view_as_real(xq_ * freqs_cis).flatten(3)
+        if xk_ is None:
+            # no keys to rotate, due to dropout
+            return xq_out.type_as(xq).to(xq.device), xk
+        # repeat freqs along seq_len dim to match k seq_len
+        if repeat_freqs_k:
+            r = xk_.shape[-2] // xq_.shape[-2]
+            if freqs_cis.is_cuda:
+                freqs_cis = freqs_cis.repeat(*([1] * (freqs_cis.ndim - 2)), r, 1)
+            else:
+                # torch.repeat on complex numbers may not be supported on non-CUDA devices
+                # (freqs_cis has 4 dims and we repeat on dim 2) so we use expand + flatten
+                freqs_cis = freqs_cis.unsqueeze(2).expand(-1, -1, r, -1, -1).flatten(2, 3)
+        xk_out = torch.view_as_real(xk_ * freqs_cis).flatten(3)
+
+    return xq_out.type_as(xq).to(xq.device), xk_out.type_as(xk).to(xk.device)

sam2/modeling/sam/__init__.py

@@ -0,0 +1,5 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.

sam2/modeling/sam/mask_decoder.py

@@ -0,0 +1,274 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.
+
+from typing import List, Optional, Tuple, Type
+
+import torch
+from torch import nn
+
+from sam2.modeling.sam2_utils import LayerNorm2d, MLP
+
+
+class MaskDecoder(nn.Module):
+
+    def __init__(
+        self,
+        *,
+        transformer_dim: int,
+        transformer: nn.Module,
+        num_multimask_outputs: int = 3,
+        activation: Type[nn.Module] = nn.GELU,
+        iou_head_depth: int = 3,
+        iou_head_hidden_dim: int = 256,
+        use_high_res_features: bool = False,
+        iou_prediction_use_sigmoid=False,
+        dynamic_multimask_via_stability=False,
+        dynamic_multimask_stability_delta=0.05,
+        dynamic_multimask_stability_thresh=0.98,
+        pred_obj_scores: bool = False,
+        pred_obj_scores_mlp: bool = False,
+        use_multimask_token_for_obj_ptr: bool = False,
+    ) -> None:
+        """
+        Predicts masks given an image and prompt embeddings, using a
+        transformer architecture.
+
+        Arguments:
+          transformer_dim (int): the channel dimension of the transformer
+          transformer (nn.Module): the transformer used to predict masks
+          num_multimask_outputs (int): the number of masks to predict
+            when disambiguating masks
+          activation (nn.Module): the type of activation to use when
+            upscaling masks
+          iou_head_depth (int): the depth of the MLP used to predict
+            mask quality
+          iou_head_hidden_dim (int): the hidden dimension of the MLP
+            used to predict mask quality
+        """
+        super().__init__()
+        self.transformer_dim = transformer_dim
+        self.transformer = transformer
+
+        self.num_multimask_outputs = num_multimask_outputs
+
+        self.iou_token = nn.Embedding(1, transformer_dim)
+        self.num_mask_tokens = num_multimask_outputs + 1
+        self.mask_tokens = nn.Embedding(self.num_mask_tokens, transformer_dim)
+
+        self.pred_obj_scores = pred_obj_scores
+        if self.pred_obj_scores:
+            self.obj_score_token = nn.Embedding(1, transformer_dim)
+        self.use_multimask_token_for_obj_ptr = use_multimask_token_for_obj_ptr
+
+        self.output_upscaling = nn.Sequential(
+            nn.ConvTranspose2d(transformer_dim, transformer_dim // 4, kernel_size=2, stride=2),
+            LayerNorm2d(transformer_dim // 4),
+            activation(),
+            nn.ConvTranspose2d(transformer_dim // 4, transformer_dim // 8, kernel_size=2, stride=2),
+            activation(),
+        )
+        self.use_high_res_features = use_high_res_features
+        if use_high_res_features:
+            self.conv_s0 = nn.Conv2d(transformer_dim, transformer_dim // 8, kernel_size=1, stride=1)
+            self.conv_s1 = nn.Conv2d(transformer_dim, transformer_dim // 4, kernel_size=1, stride=1)
+
+        self.output_hypernetworks_mlps = nn.ModuleList(
+            [MLP(transformer_dim, transformer_dim, transformer_dim // 8, 3) for i in range(self.num_mask_tokens)])
+
+        self.iou_prediction_head = MLP(
+            transformer_dim,
+            iou_head_hidden_dim,
+            self.num_mask_tokens,
+            iou_head_depth,
+            sigmoid_output=iou_prediction_use_sigmoid,
+        )
+        if self.pred_obj_scores:
+            self.pred_obj_score_head = nn.Linear(transformer_dim, 1)
+            if pred_obj_scores_mlp:
+                self.pred_obj_score_head = MLP(transformer_dim, transformer_dim, 1, 3)
+
+        # When outputting a single mask, optionally we can dynamically fall back to the best
+        # multimask output token if the single mask output token gives low stability scores.
+        self.dynamic_multimask_via_stability = dynamic_multimask_via_stability
+        self.dynamic_multimask_stability_delta = dynamic_multimask_stability_delta
+        self.dynamic_multimask_stability_thresh = dynamic_multimask_stability_thresh
+
+    def forward(
+        self,
+        image_embeddings: torch.Tensor,
+        image_pe: torch.Tensor,
+        sparse_prompt_embeddings: torch.Tensor,
+        dense_prompt_embeddings: torch.Tensor,
+        multimask_output: bool,
+        repeat_image: bool,
+        high_res_features: Optional[List[torch.Tensor]] = None,
+    ) -> Tuple[torch.Tensor, torch.Tensor]:
+        """
+        Predict masks given image and prompt embeddings.
+
+        Arguments:
+          image_embeddings (torch.Tensor): the embeddings from the image encoder
+          image_pe (torch.Tensor): positional encoding with the shape of image_embeddings
+          sparse_prompt_embeddings (torch.Tensor): the embeddings of the points and boxes
+          dense_prompt_embeddings (torch.Tensor): the embeddings of the mask inputs
+          multimask_output (bool): Whether to return multiple masks or a single
+            mask.
+
+        Returns:
+          torch.Tensor: batched predicted masks
+          torch.Tensor: batched predictions of mask quality
+          torch.Tensor: batched SAM token for mask output
+        """
+        masks, iou_pred, mask_tokens_out, object_score_logits = self.predict_masks(
+            image_embeddings=image_embeddings,
+            image_pe=image_pe,
+            sparse_prompt_embeddings=sparse_prompt_embeddings,
+            dense_prompt_embeddings=dense_prompt_embeddings,
+            repeat_image=repeat_image,
+            high_res_features=high_res_features,
+        )
+
+        # Select the correct mask or masks for output
+        if multimask_output:
+            masks = masks[:, 1:, :, :]
+            iou_pred = iou_pred[:, 1:]
+        elif self.dynamic_multimask_via_stability and not self.training:
+            masks, iou_pred = self._dynamic_multimask_via_stability(masks, iou_pred)
+        else:
+            masks = masks[:, 0:1, :, :]
+            iou_pred = iou_pred[:, 0:1]
+
+        if multimask_output and self.use_multimask_token_for_obj_ptr:
+            sam_tokens_out = mask_tokens_out[:, 1:]  # [b, 3, c] shape
+        else:
+            # Take the mask output token. Here we *always* use the token for single mask output.
+            # At test time, even if we track after 1-click (and using multimask_output=True),
+            # we still take the single mask token here. The rationale is that we always track
+            # after multiple clicks during training, so the past tokens seen during training
+            # are always the single mask token (and we'll let it be the object-memory token).
+            sam_tokens_out = mask_tokens_out[:, 0:1]  # [b, 1, c] shape
+
+        # Prepare output
+        return masks, iou_pred, sam_tokens_out, object_score_logits
+
+    def predict_masks(
+        self,
+        image_embeddings: torch.Tensor,
+        image_pe: torch.Tensor,
+        sparse_prompt_embeddings: torch.Tensor,
+        dense_prompt_embeddings: torch.Tensor,
+        repeat_image: bool,
+        high_res_features: Optional[List[torch.Tensor]] = None,
+    ) -> Tuple[torch.Tensor, torch.Tensor]:
+        """Predicts masks. See 'forward' for more details."""
+        # Concatenate output tokens
+        s = 0
+        if self.pred_obj_scores:
+            output_tokens = torch.cat(
+                [
+                    self.obj_score_token.weight,
+                    self.iou_token.weight,
+                    self.mask_tokens.weight,
+                ],
+                dim=0,
+            )
+            s = 1
+        else:
+            output_tokens = torch.cat([self.iou_token.weight, self.mask_tokens.weight], dim=0)
+        output_tokens = output_tokens.unsqueeze(0).expand(sparse_prompt_embeddings.size(0), -1, -1)
+        tokens = torch.cat((output_tokens, sparse_prompt_embeddings), dim=1)
+
+        # Expand per-image data in batch direction to be per-mask
+        if repeat_image:
+            src = torch.repeat_interleave(image_embeddings, tokens.shape[0], dim=0)
+        else:
+            assert image_embeddings.shape[0] == tokens.shape[0]
+            src = image_embeddings
+        src = src + dense_prompt_embeddings
+        assert (image_pe.size(0) == 1), "image_pe should have size 1 in batch dim (from `get_dense_pe()`)"
+        pos_src = torch.repeat_interleave(image_pe, tokens.shape[0], dim=0)
+        b, c, h, w = src.shape
+
+        # Run the transformer
+        hs, src = self.transformer(src, pos_src, tokens)
+        iou_token_out = hs[:, s, :]
+        mask_tokens_out = hs[:, s + 1:(s + 1 + self.num_mask_tokens), :]
+
+        # Upscale mask embeddings and predict masks using the mask tokens
+        src = src.transpose(1, 2).view(b, c, h, w)
+        if not self.use_high_res_features:
+            upscaled_embedding = self.output_upscaling(src)
+        else:
+            dc1, ln1, act1, dc2, act2 = self.output_upscaling
+            feat_s0, feat_s1 = high_res_features
+            upscaled_embedding = act1(ln1(dc1(src) + feat_s1))
+            upscaled_embedding = act2(dc2(upscaled_embedding) + feat_s0)
+
+        hyper_in_list: List[torch.Tensor] = []
+        for i in range(self.num_mask_tokens):
+            hyper_in_list.append(self.output_hypernetworks_mlps[i](mask_tokens_out[:, i, :]))
+        hyper_in = torch.stack(hyper_in_list, dim=1)
+        b, c, h, w = upscaled_embedding.shape
+        masks = (hyper_in @ upscaled_embedding.view(b, c, h * w)).view(b, -1, h, w)
+
+        # Generate mask quality predictions
+        iou_pred = self.iou_prediction_head(iou_token_out)
+        if self.pred_obj_scores:
+            assert s == 1
+            object_score_logits = self.pred_obj_score_head(hs[:, 0, :])
+        else:
+            # Obj scores logits - default to 10.0, i.e. assuming the object is present, sigmoid(10)=1
+            object_score_logits = 10.0 * iou_pred.new_ones(iou_pred.shape[0], 1)
+
+        return masks, iou_pred, mask_tokens_out, object_score_logits
+
+    def _get_stability_scores(self, mask_logits):
+        """
+        Compute stability scores of the mask logits based on the IoU between upper and
+        lower thresholds.
+        """
+        mask_logits = mask_logits.flatten(-2)
+        stability_delta = self.dynamic_multimask_stability_delta
+        area_i = torch.sum(mask_logits > stability_delta, dim=-1).float()
+        area_u = torch.sum(mask_logits > -stability_delta, dim=-1).float()
+        stability_scores = torch.where(area_u > 0, area_i / area_u, 1.0)
+        return stability_scores
+
+    def _dynamic_multimask_via_stability(self, all_mask_logits, all_iou_scores):
+        """
+        When outputting a single mask, if the stability score from the current single-mask
+        output (based on output token 0) falls below a threshold, we instead select from
+        multi-mask outputs (based on output token 1~3) the mask with the highest predicted
+        IoU score. This is intended to ensure a valid mask for both clicking and tracking.
+        """
+        # The best mask from multimask output tokens (1~3)
+        multimask_logits = all_mask_logits[:, 1:, :, :]
+        multimask_iou_scores = all_iou_scores[:, 1:]
+        best_scores_inds = torch.argmax(multimask_iou_scores, dim=-1)
+        batch_inds = torch.arange(multimask_iou_scores.size(0), device=all_iou_scores.device)
+        best_multimask_logits = multimask_logits[batch_inds, best_scores_inds]
+        best_multimask_logits = best_multimask_logits.unsqueeze(1)
+        best_multimask_iou_scores = multimask_iou_scores[batch_inds, best_scores_inds]
+        best_multimask_iou_scores = best_multimask_iou_scores.unsqueeze(1)
+
+        # The mask from singlemask output token 0 and its stability score
+        singlemask_logits = all_mask_logits[:, 0:1, :, :]
+        singlemask_iou_scores = all_iou_scores[:, 0:1]
+        stability_scores = self._get_stability_scores(singlemask_logits)
+        is_stable = stability_scores >= self.dynamic_multimask_stability_thresh
+
+        # Dynamically fall back to best multimask output upon low stability scores.
+        mask_logits_out = torch.where(
+            is_stable[..., None, None].expand_as(singlemask_logits),
+            singlemask_logits,
+            best_multimask_logits,
+        )
+        iou_scores_out = torch.where(
+            is_stable.expand_as(singlemask_iou_scores),
+            singlemask_iou_scores,
+            best_multimask_iou_scores,
+        )
+        return mask_logits_out, iou_scores_out

sam2/modeling/sam/prompt_encoder.py

@@ -0,0 +1,188 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.
+
+from typing import Optional, Tuple, Type
+
+import torch
+from torch import nn
+
+from sam2.modeling.position_encoding import PositionEmbeddingRandom
+from sam2.modeling.sam2_utils import LayerNorm2d
+
+
+class PromptEncoder(nn.Module):
+
+    def __init__(
+        self,
+        embed_dim: int,
+        image_embedding_size: Tuple[int, int],
+        input_image_size: Tuple[int, int],
+        mask_in_chans: int,
+        activation: Type[nn.Module] = nn.GELU,
+    ) -> None:
+        """
+        Encodes prompts for input to SAM's mask decoder.
+
+        Arguments:
+          embed_dim (int): The prompts' embedding dimension
+          image_embedding_size (tuple(int, int)): The spatial size of the
+            image embedding, as (H, W).
+          input_image_size (int): The padded size of the image as input
+            to the image encoder, as (H, W).
+          mask_in_chans (int): The number of hidden channels used for
+            encoding input masks.
+          activation (nn.Module): The activation to use when encoding
+            input masks.
+        """
+        super().__init__()
+        self.embed_dim = embed_dim
+        self.input_image_size = input_image_size
+        self.image_embedding_size = image_embedding_size
+        self.pe_layer = PositionEmbeddingRandom(embed_dim // 2)
+
+        self.num_point_embeddings: int = 4  # pos/neg point + 2 box corners
+        point_embeddings = [nn.Embedding(1, embed_dim) for i in range(self.num_point_embeddings)]
+        self.point_embeddings = nn.ModuleList(point_embeddings)
+        self.not_a_point_embed = nn.Embedding(1, embed_dim)
+
+        self.mask_input_size = (
+            4 * image_embedding_size[0],
+            4 * image_embedding_size[1],
+        )
+        self.mask_downscaling = nn.Sequential(
+            nn.Conv2d(1, mask_in_chans // 4, kernel_size=2, stride=2),
+            LayerNorm2d(mask_in_chans // 4),
+            activation(),
+            nn.Conv2d(mask_in_chans // 4, mask_in_chans, kernel_size=2, stride=2),
+            LayerNorm2d(mask_in_chans),
+            activation(),
+            nn.Conv2d(mask_in_chans, embed_dim, kernel_size=1),
+        )
+        self.no_mask_embed = nn.Embedding(1, embed_dim)
+
+    def get_dense_pe(self) -> torch.Tensor:
+        """
+        Returns the positional encoding used to encode point prompts,
+        applied to a dense set of points the shape of the image encoding.
+
+        Returns:
+          torch.Tensor: Positional encoding with shape
+            1x(embed_dim)x(embedding_h)x(embedding_w)
+        """
+        return self.pe_layer(self.image_embedding_size).unsqueeze(0)
+
+    def _embed_points(
+        self,
+        points: torch.Tensor,
+        labels: torch.Tensor,
+        pad: bool,
+    ) -> torch.Tensor:
+        """Embeds point prompts."""
+        points = points + 0.5  # Shift to center of pixel
+        if pad:
+            padding_point = torch.zeros((points.shape[0], 1, 2), device=points.device)
+            padding_label = -torch.ones((labels.shape[0], 1), device=labels.device)
+            points = torch.cat([points, padding_point], dim=1)
+            labels = torch.cat([labels, padding_label], dim=1)
+        point_embedding = self.pe_layer.forward_with_coords(points, self.input_image_size)
+        point_embedding = torch.where((labels == -1).unsqueeze(-1),
+                                      torch.zeros_like(point_embedding) + self.not_a_point_embed.weight,
+                                      point_embedding)
+        point_embedding = torch.where((labels == 0).unsqueeze(-1), point_embedding + self.point_embeddings[0].weight,
+                                      point_embedding)
+        point_embedding = torch.where((labels == 1).unsqueeze(-1), point_embedding + self.point_embeddings[1].weight,
+                                      point_embedding)
+        point_embedding = torch.where((labels == 2).unsqueeze(-1), point_embedding + self.point_embeddings[2].weight,
+                                      point_embedding)
+        point_embedding = torch.where((labels == 3).unsqueeze(-1), point_embedding + self.point_embeddings[3].weight,
+                                      point_embedding)
+        return point_embedding
+
+    def _embed_boxes(self, boxes: torch.Tensor) -> torch.Tensor:
+        """Embeds box prompts."""
+        boxes = boxes + 0.5  # Shift to center of pixel
+        coords = boxes.reshape(-1, 2, 2)
+        corner_embedding = self.pe_layer.forward_with_coords(coords, self.input_image_size)
+        corner_embedding[:, 0, :] += self.point_embeddings[2].weight
+        corner_embedding[:, 1, :] += self.point_embeddings[3].weight
+        return corner_embedding
+
+    def _embed_masks(self, masks: torch.Tensor) -> torch.Tensor:
+        """Embeds mask inputs."""
+        mask_embedding = self.mask_downscaling(masks)
+        return mask_embedding
+
+    def _get_batch_size(
+        self,
+        points: Optional[Tuple[torch.Tensor, torch.Tensor]],
+        boxes: Optional[torch.Tensor],
+        masks: Optional[torch.Tensor],
+        hidden: Optional[torch.Tensor],
+    ) -> int:
+        """
+        Gets the batch size of the output given the batch size of the input prompts.
+        """
+        if points is not None:
+            return points[0].shape[0]
+        elif boxes is not None:
+            return boxes.shape[0]
+        elif masks is not None:
+            return masks.shape[0]
+        elif hidden is not None:
+            return hidden.shape[0]
+        else:
+            return 1
+
+    def _get_device(self) -> torch.device:
+        return self.point_embeddings[0].weight.device
+
+    def forward(
+        self,
+        points: Optional[Tuple[torch.Tensor, torch.Tensor]],
+        boxes: Optional[torch.Tensor],
+        masks: Optional[torch.Tensor],
+        hidden: Optional[torch.Tensor] = None,
+    ) -> Tuple[torch.Tensor, torch.Tensor]:
+        """
+        Embeds different types of prompts, returning both sparse and dense
+        embeddings.
+
+        Arguments:
+          points (tuple(torch.Tensor, torch.Tensor) or none): point coordinates
+            and labels to embed.
+          boxes (torch.Tensor or none): boxes to embed
+          masks (torch.Tensor or none): masks to embed
+
+        Returns:
+          torch.Tensor: sparse embeddings for the points and boxes, with shape
+            BxNx(embed_dim), where N is determined by the number of input points
+            and boxes.
+          torch.Tensor: dense embeddings for the masks, in the shape
+            Bx(embed_dim)x(embed_H)x(embed_W)
+        """
+        bs = self._get_batch_size(points, boxes, masks, hidden)
+        sparse_embeddings = torch.empty((bs, 0, self.embed_dim),
+                                        dtype=self.no_mask_embed.weight.dtype,
+                                        device=self._get_device())
+        if points is not None:
+            coords, labels = points
+            point_embeddings = self._embed_points(coords, labels, pad=(boxes is None))
+            sparse_embeddings = torch.cat([sparse_embeddings, point_embeddings], dim=1)
+        if boxes is not None:
+            box_embeddings = self._embed_boxes(boxes)
+            sparse_embeddings = torch.cat([sparse_embeddings, box_embeddings], dim=1)
+
+        if hidden is not None:
+            sparse_embeddings = torch.cat([sparse_embeddings, hidden], dim=1)
+
+        if masks is not None:
+            dense_embeddings = self._embed_masks(masks)
+        else:
+            dense_embeddings = self.no_mask_embed.weight.reshape(1, -1, 1,
+                                                                 1).expand(bs, -1, self.image_embedding_size[0],
+                                                                           self.image_embedding_size[1])
+
+        return sparse_embeddings, dense_embeddings

sam2/modeling/sam/transformer.py

@@ -0,0 +1,303 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.
+
+import math
+from functools import partial
+from typing import Tuple, Type
+
+import torch
+import torch.nn.functional as F
+from torch import Tensor, nn
+
+from sam2.modeling.position_encoding import apply_rotary_enc, compute_axial_cis
+from sam2.modeling.sam2_utils import MLP
+
+
+class TwoWayTransformer(nn.Module):
+
+    def __init__(
+        self,
+        depth: int,
+        embedding_dim: int,
+        num_heads: int,
+        mlp_dim: int,
+        activation: Type[nn.Module] = nn.ReLU,
+        attention_downsample_rate: int = 2,
+    ) -> None:
+        """
+        A transformer decoder that attends to an input image using
+        queries whose positional embedding is supplied.
+
+        Args:
+          depth (int): number of layers in the transformer
+          embedding_dim (int): the channel dimension for the input embeddings
+          num_heads (int): the number of heads for multihead attention. Must
+            divide embedding_dim
+          mlp_dim (int): the channel dimension internal to the MLP block
+          activation (nn.Module): the activation to use in the MLP block
+        """
+        super().__init__()
+        self.depth = depth
+        self.embedding_dim = embedding_dim
+        self.num_heads = num_heads
+        self.mlp_dim = mlp_dim
+        self.layers = nn.ModuleList()
+
+        for i in range(depth):
+            self.layers.append(
+                TwoWayAttentionBlock(
+                    embedding_dim=embedding_dim,
+                    num_heads=num_heads,
+                    mlp_dim=mlp_dim,
+                    activation=activation,
+                    attention_downsample_rate=attention_downsample_rate,
+                    skip_first_layer_pe=(i == 0),
+                ))
+
+        self.final_attn_token_to_image = Attention(embedding_dim, num_heads, downsample_rate=attention_downsample_rate)
+        self.norm_final_attn = nn.LayerNorm(embedding_dim)
+
+    def forward(
+        self,
+        image_embedding: Tensor,
+        image_pe: Tensor,
+        point_embedding: Tensor,
+    ) -> Tuple[Tensor, Tensor]:
+        """
+        Args:
+          image_embedding (torch.Tensor): image to attend to. Should be shape
+            B x embedding_dim x h x w for any h and w.
+          image_pe (torch.Tensor): the positional encoding to add to the image. Must
+            have the same shape as image_embedding.
+          point_embedding (torch.Tensor): the embedding to add to the query points.
+            Must have shape B x N_points x embedding_dim for any N_points.
+
+        Returns:
+          torch.Tensor: the processed point_embedding
+          torch.Tensor: the processed image_embedding
+        """
+        # BxCxHxW -> BxHWxC == B x N_image_tokens x C
+        bs, c, h, w = image_embedding.shape
+        image_embedding = image_embedding.flatten(2).permute(0, 2, 1)
+        image_pe = image_pe.flatten(2).permute(0, 2, 1)
+
+        # Prepare queries
+        queries = point_embedding
+        keys = image_embedding
+
+        # Apply transformer blocks and final layernorm
+        for layer in self.layers:
+            queries, keys = layer(
+                queries=queries,
+                keys=keys,
+                query_pe=point_embedding,
+                key_pe=image_pe,
+            )
+
+        # Apply the final attention layer from the points to the image
+        q = queries + point_embedding
+        k = keys + image_pe
+        attn_out = self.final_attn_token_to_image(q=q, k=k, v=keys)
+        queries = queries + attn_out
+        queries = self.norm_final_attn(queries)
+
+        return queries, keys
+
+
+class TwoWayAttentionBlock(nn.Module):
+
+    def __init__(
+        self,
+        embedding_dim: int,
+        num_heads: int,
+        mlp_dim: int = 2048,
+        activation: Type[nn.Module] = nn.ReLU,
+        attention_downsample_rate: int = 2,
+        skip_first_layer_pe: bool = False,
+    ) -> None:
+        """
+        A transformer block with four layers: (1) self-attention of sparse
+        inputs, (2) cross attention of sparse inputs to dense inputs, (3) mlp
+        block on sparse inputs, and (4) cross attention of dense inputs to sparse
+        inputs.
+
+        Arguments:
+          embedding_dim (int): the channel dimension of the embeddings
+          num_heads (int): the number of heads in the attention layers
+          mlp_dim (int): the hidden dimension of the mlp block
+          activation (nn.Module): the activation of the mlp block
+          skip_first_layer_pe (bool): skip the PE on the first layer
+        """
+        super().__init__()
+        self.self_attn = Attention(embedding_dim, num_heads)
+        self.norm1 = nn.LayerNorm(embedding_dim)
+
+        self.cross_attn_token_to_image = Attention(embedding_dim, num_heads, downsample_rate=attention_downsample_rate)
+        self.norm2 = nn.LayerNorm(embedding_dim)
+
+        self.mlp = MLP(embedding_dim, mlp_dim, embedding_dim, num_layers=2, activation=activation)
+        self.norm3 = nn.LayerNorm(embedding_dim)
+
+        self.norm4 = nn.LayerNorm(embedding_dim)
+        self.cross_attn_image_to_token = Attention(embedding_dim, num_heads, downsample_rate=attention_downsample_rate)
+
+        self.skip_first_layer_pe = skip_first_layer_pe
+
+    def forward(self, queries: Tensor, keys: Tensor, query_pe: Tensor, key_pe: Tensor) -> Tuple[Tensor, Tensor]:
+        # Self attention block
+        if self.skip_first_layer_pe:
+            queries = self.self_attn(q=queries, k=queries, v=queries)
+        else:
+            q = queries + query_pe
+            attn_out = self.self_attn(q=q, k=q, v=queries)
+            queries = queries + attn_out
+        queries = self.norm1(queries)
+
+        # Cross attention block, tokens attending to image embedding
+        q = queries + query_pe
+        k = keys + key_pe
+        attn_out = self.cross_attn_token_to_image(q=q, k=k, v=keys)
+        queries = queries + attn_out
+        queries = self.norm2(queries)
+
+        # MLP block
+        mlp_out = self.mlp(queries)
+        queries = queries + mlp_out
+        queries = self.norm3(queries)
+
+        # Cross attention block, image embedding attending to tokens
+        q = queries + query_pe
+        k = keys + key_pe
+        attn_out = self.cross_attn_image_to_token(q=k, k=q, v=queries)
+        keys = keys + attn_out
+        keys = self.norm4(keys)
+
+        return queries, keys
+
+
+class Attention(nn.Module):
+    """
+    An attention layer that allows for downscaling the size of the embedding
+    after projection to queries, keys, and values.
+    """
+
+    def __init__(
+        self,
+        embedding_dim: int,
+        num_heads: int,
+        downsample_rate: int = 1,
+        dropout: float = 0.0,
+        kv_in_dim: int = None,
+    ) -> None:
+        super().__init__()
+        self.embedding_dim = embedding_dim
+        self.kv_in_dim = kv_in_dim if kv_in_dim is not None else embedding_dim
+        self.internal_dim = embedding_dim // downsample_rate
+        self.num_heads = num_heads
+        assert (self.internal_dim % num_heads == 0), "num_heads must divide embedding_dim."
+
+        self.q_proj = nn.Linear(embedding_dim, self.internal_dim)
+        self.k_proj = nn.Linear(self.kv_in_dim, self.internal_dim)
+        self.v_proj = nn.Linear(self.kv_in_dim, self.internal_dim)
+        self.out_proj = nn.Linear(self.internal_dim, embedding_dim)
+
+        self.dropout_p = dropout
+
+    def _separate_heads(self, x: Tensor, num_heads: int) -> Tensor:
+        b, n, c = x.shape
+        x = x.reshape(b, n, num_heads, c // num_heads)
+        return x.transpose(1, 2)  # B x N_heads x N_tokens x C_per_head
+
+    def _recombine_heads(self, x: Tensor) -> Tensor:
+        b, n_heads, n_tokens, c_per_head = x.shape
+        x = x.transpose(1, 2)
+        return x.reshape(b, n_tokens, n_heads * c_per_head)  # B x N_tokens x C
+
+    def forward(self, q: Tensor, k: Tensor, v: Tensor) -> Tensor:
+        # Input projections
+        q = self.q_proj(q)
+        k = self.k_proj(k)
+        v = self.v_proj(v)
+
+        # Separate into heads
+        q = self._separate_heads(q, self.num_heads)
+        k = self._separate_heads(k, self.num_heads)
+        v = self._separate_heads(v, self.num_heads)
+
+        dropout_p = self.dropout_p if self.training else 0.0
+        # Attention
+        out = F.scaled_dot_product_attention(q, k, v, dropout_p=dropout_p)
+
+        out = self._recombine_heads(out)
+        out = self.out_proj(out)
+
+        return out
+
+
+class RoPEAttention(Attention):
+    """Attention with rotary position encoding."""
+
+    def __init__(
+            self,
+            *args,
+            rope_theta=10000.0,
+            # whether to repeat q rope to match k length
+            # this is needed for cross-attention to memories
+            rope_k_repeat=False,
+            feat_sizes=(64, 64),  # [w, h] for stride 16 feats at 1024 resolution
+            **kwargs,
+    ):
+        super().__init__(*args, **kwargs)
+
+        self.compute_cis = partial(compute_axial_cis, dim=self.internal_dim // self.num_heads, theta=rope_theta)
+        freqs_cis = self.compute_cis(end_x=feat_sizes[0], end_y=feat_sizes[1])
+        try:
+            import torch_npu
+            has_npu = torch_npu.npu.is_available()
+        except ImportError:
+            has_npu = False
+        if torch.cuda.is_available():
+            freqs_cis = freqs_cis.to("cuda")
+        elif has_npu:
+            freqs_cis = freqs_cis.to("npu")
+        self.freqs_cis = freqs_cis
+        self.rope_k_repeat = rope_k_repeat
+
+    def forward(self, q: Tensor, k: Tensor, v: Tensor, num_k_exclude_rope: int = 0) -> Tensor:
+        # Input projections
+        q = self.q_proj(q)
+        k = self.k_proj(k)
+        v = self.v_proj(v)
+
+        # Separate into heads
+        q = self._separate_heads(q, self.num_heads)
+        k = self._separate_heads(k, self.num_heads)
+        v = self._separate_heads(v, self.num_heads)
+
+        # Apply rotary position encoding
+        w = h = math.sqrt(q.shape[-2])
+        self.freqs_cis = self.freqs_cis.to(q.device)
+        if self.freqs_cis.shape[0] != q.shape[-2]:
+            self.freqs_cis = self.compute_cis(end_x=w, end_y=h).to(q.device)
+        if q.shape[-2] != k.shape[-2]:
+            assert self.rope_k_repeat
+
+        num_k_rope = k.size(-2) - num_k_exclude_rope
+        q, k[:, :, :num_k_rope] = apply_rotary_enc(
+            q,
+            k[:, :, :num_k_rope],
+            freqs_cis=self.freqs_cis,
+            repeat_freqs_k=self.rope_k_repeat,
+        )
+
+        dropout_p = self.dropout_p if self.training else 0.0
+        # Attention
+        out = F.scaled_dot_product_attention(q, k, v, dropout_p=dropout_p)
+
+        out = self._recombine_heads(out)
+        out = self.out_proj(out)
+
+        return out

sam2/modeling/sam2_base.py

@@ -0,0 +1,882 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.
+
+import torch
+import torch.distributed
+import torch.nn.functional as F
+from torch.nn.init import trunc_normal_
+
+from sam2.modeling.sam2_utils import MLP, get_1d_sine_pe, select_closest_cond_frames
+from sam2.modeling.sam.mask_decoder import MaskDecoder
+from sam2.modeling.sam.prompt_encoder import PromptEncoder
+from sam2.modeling.sam.transformer import TwoWayTransformer
+
+# a large negative value as a placeholder score for missing objects
+NO_OBJ_SCORE = -1024.0
+
+
+class SAM2Base(torch.nn.Module):
+
+    def __init__(
+        self,
+        image_encoder,
+        memory_attention,
+        memory_encoder,
+        num_maskmem=7,  # default 1 input frame + 6 previous frames
+        image_size=512,
+        backbone_stride=16,  # stride of the image backbone output
+        sigmoid_scale_for_mem_enc=1.0,  # scale factor for mask sigmoid prob
+        sigmoid_bias_for_mem_enc=0.0,  # bias factor for mask sigmoid prob
+        # During evaluation, whether to binarize the sigmoid mask logits on interacted frames with clicks
+        binarize_mask_from_pts_for_mem_enc=False,
+        use_mask_input_as_output_without_sam=False,  # on frames with mask input, whether to directly output the input mask without using a SAM prompt encoder + mask decoder
+        # The maximum number of conditioning frames to participate in the memory attention (-1 means no limit; if there are more conditioning frames than this limit,
+        # we only cross-attend to the temporally closest `max_cond_frames_in_attn` conditioning frames in the encoder when tracking each frame). This gives the model
+        # a temporal locality when handling a large number of annotated frames (since closer frames should be more important) and also avoids GPU OOM.
+        max_cond_frames_in_attn=-1,
+        # on the first frame, whether to directly add the no-memory embedding to the image feature
+        # (instead of using the transformer encoder)
+        directly_add_no_mem_embed=False,
+        # whether to use high-resolution feature maps in the SAM mask decoder
+        use_high_res_features_in_sam=False,
+        # whether to output multiple (3) masks for the first click on initial conditioning frames
+        multimask_output_in_sam=False,
+        # the minimum and maximum number of clicks to use multimask_output_in_sam (only relevant when `multimask_output_in_sam=True`;
+        # default is 1 for both, meaning that only the first click gives multimask output; also note that a box counts as two points)
+        multimask_min_pt_num=1,
+        multimask_max_pt_num=1,
+        # whether to also use multimask output for tracking (not just for the first click on initial conditioning frames; only relevant when `multimask_output_in_sam=True`)
+        multimask_output_for_tracking=False,
+        # Whether to use multimask tokens for obj ptr; Only relevant when both
+        # use_obj_ptrs_in_encoder=True and multimask_output_for_tracking=True
+        use_multimask_token_for_obj_ptr: bool = False,
+        # whether to use sigmoid to restrict ious prediction to [0-1]
+        iou_prediction_use_sigmoid=False,
+        # The memory bank's temporal stride during evaluation (i.e. the `r` parameter in XMem and Cutie; XMem and Cutie use r=5).
+        # For r>1, the (self.num_maskmem - 1) non-conditioning memory frames consist of
+        # (self.num_maskmem - 2) nearest frames from every r-th frames, plus the last frame.
+        memory_temporal_stride_for_eval=1,
+        # whether to apply non-overlapping constraints on the object masks in the memory encoder during evaluation (to avoid/alleviate superposing masks)
+        non_overlap_masks_for_mem_enc=False,
+        # whether to cross-attend to object pointers from other frames (based on SAM output tokens) in the encoder
+        use_obj_ptrs_in_encoder=False,
+        # the maximum number of object pointers from other frames in encoder cross attention (only relevant when `use_obj_ptrs_in_encoder=True`)
+        max_obj_ptrs_in_encoder=16,
+        # whether to add temporal positional encoding to the object pointers in the encoder (only relevant when `use_obj_ptrs_in_encoder=True`)
+        add_tpos_enc_to_obj_ptrs=True,
+        # whether to add an extra linear projection layer for the temporal positional encoding in the object pointers to avoid potential interference
+        # with spatial positional encoding (only relevant when both `use_obj_ptrs_in_encoder=True` and `add_tpos_enc_to_obj_ptrs=True`)
+        proj_tpos_enc_in_obj_ptrs=False,
+        # whether to use signed distance (instead of unsigned absolute distance) in the temporal positional encoding in the object pointers
+        # (only relevant when both `use_obj_ptrs_in_encoder=True` and `add_tpos_enc_to_obj_ptrs=True`)
+        use_signed_tpos_enc_to_obj_ptrs=False,
+        # whether to only attend to object pointers in the past (before the current frame) in the encoder during evaluation
+        # (only relevant when `use_obj_ptrs_in_encoder=True`; this might avoid pointer information too far in the future to distract the initial tracking)
+        only_obj_ptrs_in_the_past_for_eval=False,
+        # Whether to predict if there is an object in the frame
+        pred_obj_scores: bool = False,
+        # Whether to use an MLP to predict object scores
+        pred_obj_scores_mlp: bool = False,
+        # Only relevant if pred_obj_scores=True and use_obj_ptrs_in_encoder=True;
+        # Whether to have a fixed no obj pointer when there is no object present
+        # or to use it as an additive embedding with obj_ptr produced by decoder
+        fixed_no_obj_ptr: bool = False,
+        # Soft no object, i.e. mix in no_obj_ptr softly,
+        # hope to make recovery easier if there is a mistake and mitigate accumulation of errors
+        soft_no_obj_ptr: bool = False,
+        use_mlp_for_obj_ptr_proj: bool = False,
+        # add no obj embedding to spatial frames
+        no_obj_embed_spatial: bool = False,
+        # extra arguments used to construct the SAM mask decoder; if not None, it should be a dict of kwargs to be passed into `MaskDecoder` class.
+        sam_mask_decoder_extra_args=None,
+        compile_image_encoder: bool = False,
+        **kwargs,
+    ):
+        super().__init__()
+
+        # Part 1: the image backbone
+        self.image_encoder = image_encoder
+        # Use level 0, 1, 2 for high-res setting, or just level 2 for the default setting
+        self.use_high_res_features_in_sam = use_high_res_features_in_sam
+        self.num_feature_levels = 3 if use_high_res_features_in_sam else 1
+        self.use_obj_ptrs_in_encoder = use_obj_ptrs_in_encoder
+        self.max_obj_ptrs_in_encoder = max_obj_ptrs_in_encoder
+        if use_obj_ptrs_in_encoder:
+            # A conv layer to downsample the mask prompt to stride 4 (the same stride as
+            # low-res SAM mask logits) and to change its scales from 0~1 to SAM logit scale,
+            # so that it can be fed into the SAM mask decoder to generate a pointer.
+            self.mask_downsample = torch.nn.Conv2d(1, 1, kernel_size=4, stride=4)
+        self.add_tpos_enc_to_obj_ptrs = add_tpos_enc_to_obj_ptrs
+        if proj_tpos_enc_in_obj_ptrs:
+            assert add_tpos_enc_to_obj_ptrs  # these options need to be used together
+        self.proj_tpos_enc_in_obj_ptrs = proj_tpos_enc_in_obj_ptrs
+        self.use_signed_tpos_enc_to_obj_ptrs = use_signed_tpos_enc_to_obj_ptrs
+        self.only_obj_ptrs_in_the_past_for_eval = only_obj_ptrs_in_the_past_for_eval
+
+        # Part 2: memory attention to condition current frame's visual features
+        # with memories (and obj ptrs) from past frames
+        self.memory_attention = memory_attention
+        self.hidden_dim = image_encoder.neck.d_model
+
+        # Part 3: memory encoder for the previous frame's outputs
+        self.memory_encoder = memory_encoder
+        self.mem_dim = self.hidden_dim
+        if hasattr(self.memory_encoder, "out_proj") and hasattr(self.memory_encoder.out_proj, "weight"):
+            # if there is compression of memories along channel dim
+            self.mem_dim = self.memory_encoder.out_proj.weight.shape[0]
+        self.num_maskmem = num_maskmem  # Number of memories accessible
+        # Temporal encoding of the memories
+        self.maskmem_tpos_enc = torch.nn.Parameter(torch.zeros(num_maskmem, 1, 1, self.mem_dim))
+        trunc_normal_(self.maskmem_tpos_enc, std=0.02)
+        # a single token to indicate no memory embedding from previous frames
+        self.no_mem_embed = torch.nn.Parameter(torch.zeros(1, 1, self.hidden_dim))
+        self.no_mem_pos_enc = torch.nn.Parameter(torch.zeros(1, 1, self.hidden_dim))
+        trunc_normal_(self.no_mem_embed, std=0.02)
+        trunc_normal_(self.no_mem_pos_enc, std=0.02)
+        self.directly_add_no_mem_embed = directly_add_no_mem_embed
+        # Apply sigmoid to the output raw mask logits (to turn them from
+        # range (-inf, +inf) to range (0, 1)) before feeding them into the memory encoder
+        self.sigmoid_scale_for_mem_enc = sigmoid_scale_for_mem_enc
+        self.sigmoid_bias_for_mem_enc = sigmoid_bias_for_mem_enc
+        self.binarize_mask_from_pts_for_mem_enc = binarize_mask_from_pts_for_mem_enc
+        self.non_overlap_masks_for_mem_enc = non_overlap_masks_for_mem_enc
+        self.memory_temporal_stride_for_eval = memory_temporal_stride_for_eval
+        # On frames with mask input, whether to directly output the input mask without
+        # using a SAM prompt encoder + mask decoder
+        self.use_mask_input_as_output_without_sam = use_mask_input_as_output_without_sam
+        self.multimask_output_in_sam = multimask_output_in_sam
+        self.multimask_min_pt_num = multimask_min_pt_num
+        self.multimask_max_pt_num = multimask_max_pt_num
+        self.multimask_output_for_tracking = multimask_output_for_tracking
+        self.use_multimask_token_for_obj_ptr = use_multimask_token_for_obj_ptr
+        self.iou_prediction_use_sigmoid = iou_prediction_use_sigmoid
+
+        # Part 4: SAM-style prompt encoder (for both mask and point inputs)
+        # and SAM-style mask decoder for the final mask output
+        self.image_size = image_size
+        self.backbone_stride = backbone_stride
+        self.sam_mask_decoder_extra_args = sam_mask_decoder_extra_args
+        self.pred_obj_scores = pred_obj_scores
+        self.pred_obj_scores_mlp = pred_obj_scores_mlp
+        self.fixed_no_obj_ptr = fixed_no_obj_ptr
+        self.soft_no_obj_ptr = soft_no_obj_ptr
+        if self.fixed_no_obj_ptr:
+            assert self.pred_obj_scores
+            assert self.use_obj_ptrs_in_encoder
+        if self.pred_obj_scores and self.use_obj_ptrs_in_encoder:
+            self.no_obj_ptr = torch.nn.Parameter(torch.zeros(1, self.hidden_dim))
+            trunc_normal_(self.no_obj_ptr, std=0.02)
+        self.use_mlp_for_obj_ptr_proj = use_mlp_for_obj_ptr_proj
+        self.no_obj_embed_spatial = None
+        if no_obj_embed_spatial:
+            self.no_obj_embed_spatial = torch.nn.Parameter(torch.zeros(1, self.mem_dim))
+            trunc_normal_(self.no_obj_embed_spatial, std=0.02)
+
+        self._build_sam_heads()
+        self.max_cond_frames_in_attn = max_cond_frames_in_attn
+
+        # Model compilation
+        if compile_image_encoder:
+            # Compile the forward function (not the full module) to allow loading checkpoints.
+            print("Image encoder compilation is enabled. First forward pass will be slow.")
+            self.image_encoder.forward = torch.compile(
+                self.image_encoder.forward,
+                mode="max-autotune",
+                fullgraph=True,
+                dynamic=False,
+            )
+
+    @property
+    def device(self):
+        return next(self.parameters()).device
+
+    def forward(self, *args, **kwargs):
+        raise NotImplementedError(
+            "Please use the corresponding methods in SAM2VideoPredictor for inference or SAM2Train for training/fine-tuning"
+            "See notebooks/video_predictor_example.ipynb for an inference example.")
+
+    def _build_sam_heads(self):
+        """Build SAM-style prompt encoder and mask decoder."""
+        self.sam_prompt_embed_dim = self.hidden_dim
+        self.sam_image_embedding_size = self.image_size // self.backbone_stride
+
+        # build PromptEncoder and MaskDecoder from SAM
+        # (their hyperparameters like `mask_in_chans=16` are from SAM code)
+        self.sam_prompt_encoder = PromptEncoder(
+            embed_dim=self.sam_prompt_embed_dim,
+            image_embedding_size=(
+                self.sam_image_embedding_size,
+                self.sam_image_embedding_size,
+            ),
+            input_image_size=(self.image_size, self.image_size),
+            mask_in_chans=16,
+        )
+        self.sam_mask_decoder = MaskDecoder(
+            num_multimask_outputs=3,
+            transformer=TwoWayTransformer(
+                depth=2,
+                embedding_dim=self.sam_prompt_embed_dim,
+                mlp_dim=2048,
+                num_heads=8,
+            ),
+            transformer_dim=self.sam_prompt_embed_dim,
+            iou_head_depth=3,
+            iou_head_hidden_dim=256,
+            use_high_res_features=self.use_high_res_features_in_sam,
+            iou_prediction_use_sigmoid=self.iou_prediction_use_sigmoid,
+            pred_obj_scores=self.pred_obj_scores,
+            pred_obj_scores_mlp=self.pred_obj_scores_mlp,
+            use_multimask_token_for_obj_ptr=self.use_multimask_token_for_obj_ptr,
+            **(self.sam_mask_decoder_extra_args or {}),
+        )
+        if self.use_obj_ptrs_in_encoder:
+            # a linear projection on SAM output tokens to turn them into object pointers
+            self.obj_ptr_proj = torch.nn.Linear(self.hidden_dim, self.hidden_dim)
+            if self.use_mlp_for_obj_ptr_proj:
+                self.obj_ptr_proj = MLP(self.hidden_dim, self.hidden_dim, self.hidden_dim, 3)
+        else:
+            self.obj_ptr_proj = torch.nn.Identity()
+        if self.proj_tpos_enc_in_obj_ptrs:
+            # a linear projection on temporal positional encoding in object pointers to
+            # avoid potential interference with spatial positional encoding
+            self.obj_ptr_tpos_proj = torch.nn.Linear(self.hidden_dim, self.mem_dim)
+        else:
+            self.obj_ptr_tpos_proj = torch.nn.Identity()
+
+    def _forward_sam_heads(
+        self,
+        backbone_features,
+        point_inputs=None,
+        mask_inputs=None,
+        hidden_inputs=None,
+        high_res_features=None,
+        multimask_output=False,
+    ):
+        """
+        Forward SAM prompt encoders and mask heads.
+
+        Inputs:
+        - backbone_features: image features of [B, C, H, W] shape
+        - point_inputs: a dictionary with "point_coords" and "point_labels", where
+          1) "point_coords" has [B, P, 2] shape and float32 dtype and contains the
+             absolute pixel-unit coordinate in (x, y) format of the P input points
+          2) "point_labels" has shape [B, P] and int32 dtype, where 1 means
+             positive clicks, 0 means negative clicks, and -1 means padding
+        - mask_inputs: a mask of [B, 1, H*16, W*16] shape, float or bool, with the
+          same spatial size as the image.
+        - high_res_features: either 1) None or 2) or a list of length 2 containing
+          two feature maps of [B, C, 4*H, 4*W] and [B, C, 2*H, 2*W] shapes respectively,
+          which will be used as high-resolution feature maps for SAM decoder.
+        - multimask_output: if it's True, we output 3 candidate masks and their 3
+          corresponding IoU estimates, and if it's False, we output only 1 mask and
+          its corresponding IoU estimate.
+
+        Outputs:
+        - low_res_multimasks: [B, M, H*4, W*4] shape (where M = 3 if
+          `multimask_output=True` and M = 1 if `multimask_output=False`), the SAM
+          output mask logits (before sigmoid) for the low-resolution masks, with 4x
+          the resolution (1/4 stride) of the input backbone_features.
+        - high_res_multimasks: [B, M, H*16, W*16] shape (where M = 3
+          if `multimask_output=True` and M = 1 if `multimask_output=False`),
+          upsampled from the low-resolution masks, with shape size as the image
+          (stride is 1 pixel).
+        - ious, [B, M] shape, where (where M = 3 if `multimask_output=True` and M = 1
+          if `multimask_output=False`), the estimated IoU of each output mask.
+        - low_res_masks: [B, 1, H*4, W*4] shape, the best mask in `low_res_multimasks`.
+          If `multimask_output=True`, it's the mask with the highest IoU estimate.
+          If `multimask_output=False`, it's the same as `low_res_multimasks`.
+        - high_res_masks: [B, 1, H*16, W*16] shape, the best mask in `high_res_multimasks`.
+          If `multimask_output=True`, it's the mask with the highest IoU estimate.
+          If `multimask_output=False`, it's the same as `high_res_multimasks`.
+        - obj_ptr: [B, C] shape, the object pointer vector for the output mask, extracted
+          based on the output token from the SAM mask decoder.
+        """
+        B = backbone_features.size(0)
+        device = backbone_features.device
+        assert backbone_features.size(1) == self.sam_prompt_embed_dim
+        assert backbone_features.size(2) == self.sam_image_embedding_size
+        assert backbone_features.size(3) == self.sam_image_embedding_size
+
+        # a) Handle point prompts
+        if point_inputs is not None:
+            sam_point_coords = point_inputs["point_coords"]
+            sam_point_labels = point_inputs["point_labels"]
+            assert sam_point_coords.size(0) == B and sam_point_labels.size(0) == B
+        else:
+            # If no points are provide, pad with an empty point (with label -1)
+            sam_point_coords = torch.zeros(B, 1, 2, device=device)
+            sam_point_labels = -torch.ones(B, 1, dtype=torch.int32, device=device)
+
+        # b) Handle mask prompts
+        if mask_inputs is not None:
+            # If mask_inputs is provided, downsize it into low-res mask input if needed
+            # and feed it as a dense mask prompt into the SAM mask encoder
+            assert len(mask_inputs.shape) == 4 and mask_inputs.shape[:2] == (B, 1)
+            if mask_inputs.shape[-2:] != self.sam_prompt_encoder.mask_input_size:
+                sam_mask_prompt = F.interpolate(
+                    mask_inputs.float(),
+                    size=self.sam_prompt_encoder.mask_input_size,
+                    align_corners=False,
+                    mode="bilinear",
+                    antialias=True,  # use antialias for downsampling
+                )
+            else:
+                sam_mask_prompt = mask_inputs
+        else:
+            # Otherwise, simply feed None (and SAM's prompt encoder will add
+            # a learned `no_mask_embed` to indicate no mask input in this case).
+            sam_mask_prompt = None
+
+        sparse_embeddings, dense_embeddings = self.sam_prompt_encoder(
+            points=(sam_point_coords, sam_point_labels),
+            boxes=None,
+            masks=sam_mask_prompt,
+            hidden=hidden_inputs,
+        )
+        (
+            low_res_multimasks,
+            ious,
+            sam_output_tokens,
+            object_score_logits,
+        ) = self.sam_mask_decoder(
+            image_embeddings=backbone_features,
+            image_pe=self.sam_prompt_encoder.get_dense_pe(),
+            sparse_prompt_embeddings=sparse_embeddings,
+            dense_prompt_embeddings=dense_embeddings,
+            multimask_output=multimask_output,
+            repeat_image=False,  # the image is already batched
+            high_res_features=high_res_features,
+        )
+        if self.pred_obj_scores:
+            is_obj_appearing = object_score_logits > 0
+
+            # Mask used for spatial memories is always a *hard* choice between obj and no obj,
+            # consistent with the actual mask prediction
+            # NOTE: whether to mask here during inference?
+            if getattr(self, 'inference_mode', False):
+                low_res_multimasks = torch.where(
+                    is_obj_appearing[:, None, None],
+                    low_res_multimasks,
+                    NO_OBJ_SCORE,
+                )
+
+        # convert masks from possibly bfloat16 (or float16) to float32
+        # low_res_multimasks = low_res_multimasks.float()
+        high_res_multimasks = F.interpolate(
+            low_res_multimasks.float(),
+            size=(self.image_size, self.image_size),
+            mode="bilinear",
+            align_corners=False,
+        ).to(low_res_multimasks.dtype)
+
+        sam_output_token = sam_output_tokens[:, 0]
+        if multimask_output:
+            # take the best mask prediction (with the highest IoU estimation)
+            best_iou_inds = torch.argmax(ious, dim=-1)
+            batch_inds = torch.arange(B, device=device)
+            low_res_masks = low_res_multimasks[batch_inds, best_iou_inds].unsqueeze(1)
+            high_res_masks = high_res_multimasks[batch_inds, best_iou_inds].unsqueeze(1)
+            if sam_output_tokens.size(1) > 1:
+                sam_output_token = sam_output_tokens[batch_inds, best_iou_inds]
+        else:
+            low_res_masks, high_res_masks = low_res_multimasks, high_res_multimasks
+
+        # Extract object pointer from the SAM output token (with occlusion handling)
+        obj_ptr = self.obj_ptr_proj(sam_output_token)
+        if self.pred_obj_scores:
+            # Allow *soft* no obj ptr, unlike for masks
+            if self.soft_no_obj_ptr:
+                lambda_is_obj_appearing = object_score_logits.sigmoid()
+            else:
+                lambda_is_obj_appearing = is_obj_appearing.to(object_score_logits.dtype)
+
+            if self.fixed_no_obj_ptr:
+                obj_ptr = lambda_is_obj_appearing * obj_ptr
+            obj_ptr = obj_ptr + (1 - lambda_is_obj_appearing) * self.no_obj_ptr
+
+        return (
+            low_res_multimasks,
+            high_res_multimasks,
+            ious,
+            low_res_masks,
+            high_res_masks,
+            obj_ptr,
+            object_score_logits,
+        )
+
+    def _use_mask_as_output(self, backbone_features, high_res_features, mask_inputs):
+        """
+        Directly turn binary `mask_inputs` into a output mask logits without using SAM.
+        (same input and output shapes as in _forward_sam_heads above).
+        """
+        # Use -10/+10 as logits for neg/pos pixels (very close to 0/1 in prob after sigmoid).
+        out_scale, out_bias = 20.0, -10.0  # sigmoid(-10.0)=4.5398e-05
+        mask_inputs_float = mask_inputs.float()
+        high_res_masks = mask_inputs_float * out_scale + out_bias
+        low_res_masks = F.interpolate(
+            high_res_masks,
+            size=(high_res_masks.size(-2) // 4, high_res_masks.size(-1) // 4),
+            align_corners=False,
+            mode="bilinear",
+            antialias=True,  # use antialias for downsampling
+        )
+        # a dummy IoU prediction of all 1's under mask input
+        ious = mask_inputs.new_ones(mask_inputs.size(0), 1).float()
+        if not self.use_obj_ptrs_in_encoder:
+            # all zeros as a dummy object pointer (of shape [B, C])
+            obj_ptr = torch.zeros(mask_inputs.size(0), self.hidden_dim, device=mask_inputs.device)
+        else:
+            # produce an object pointer using the SAM decoder from the mask input
+            _, _, _, _, _, obj_ptr, _ = self._forward_sam_heads(
+                backbone_features=backbone_features,
+                mask_inputs=self.mask_downsample(mask_inputs_float),
+                high_res_features=high_res_features,
+            )
+        # In this method, we are treating mask_input as output, e.g. using it directly to create spatial mem;
+        # Below, we follow the same design axiom to use mask_input to decide if obj appears or not instead of relying
+        # on the object_scores from the SAM decoder.
+        is_obj_appearing = torch.any(mask_inputs.flatten(1).float() > 0.0, dim=1)
+        is_obj_appearing = is_obj_appearing[..., None]
+        lambda_is_obj_appearing = is_obj_appearing.float()
+        object_score_logits = out_scale * lambda_is_obj_appearing + out_bias
+        if self.pred_obj_scores:
+            if self.fixed_no_obj_ptr:
+                obj_ptr = lambda_is_obj_appearing * obj_ptr
+            obj_ptr = obj_ptr + (1 - lambda_is_obj_appearing) * self.no_obj_ptr
+
+        return (
+            low_res_masks,
+            high_res_masks,
+            ious,
+            low_res_masks,
+            high_res_masks,
+            obj_ptr,
+            object_score_logits,
+        )
+
+    def forward_image(self, img_batch: torch.Tensor):
+        """Get the image feature on the input batch."""
+        backbone_out = self.image_encoder(img_batch)
+        if self.use_high_res_features_in_sam:
+            # precompute projected level 0 and level 1 features in SAM decoder
+            # to avoid running it again on every SAM click
+            backbone_out["backbone_fpn"][0] = self.sam_mask_decoder.conv_s0(backbone_out["backbone_fpn"][0])
+            backbone_out["backbone_fpn"][1] = self.sam_mask_decoder.conv_s1(backbone_out["backbone_fpn"][1])
+        return backbone_out
+
+    def _prepare_backbone_features(self, backbone_out):
+        """Prepare and flatten visual features."""
+        backbone_out = backbone_out.copy()
+        assert len(backbone_out["backbone_fpn"]) == len(backbone_out["vision_pos_enc"])
+        assert len(backbone_out["backbone_fpn"]) >= self.num_feature_levels
+
+        feature_maps = backbone_out["backbone_fpn"][-self.num_feature_levels:]
+        vision_pos_embeds = backbone_out["vision_pos_enc"][-self.num_feature_levels:]
+
+        feat_sizes = [(x.shape[-2], x.shape[-1]) for x in vision_pos_embeds]
+        # flatten NxCxHxW to HWxNxC
+        vision_feats = [x.flatten(2).permute(2, 0, 1) for x in feature_maps]
+        vision_pos_embeds = [x.flatten(2).permute(2, 0, 1) for x in vision_pos_embeds]
+
+        return backbone_out, vision_feats, vision_pos_embeds, feat_sizes
+
+    def _prepare_memory_conditioned_features(
+            self,
+            frame_idx,
+            is_init_cond_frame,
+            current_vision_feats,
+            current_vision_pos_embeds,
+            feat_sizes,
+            output_dict,
+            num_frames,
+            track_in_reverse=False,  # tracking in reverse time order (for demo usage)
+    ):
+        """Fuse the current frame's visual feature map with previous memory."""
+        B = current_vision_feats[-1].size(1)  # batch size on this frame
+        C = self.hidden_dim
+        H, W = feat_sizes[-1]  # top-level (lowest-resolution) feature size
+        device = current_vision_feats[-1].device
+        # The case of `self.num_maskmem == 0` below is primarily used for reproducing SAM on images.
+        # In this case, we skip the fusion with any memory.
+        if self.num_maskmem == 0:  # Disable memory and skip fusion
+            pix_feat = current_vision_feats[-1].permute(1, 2, 0).view(B, C, H, W)
+            return pix_feat
+
+        num_obj_ptr_tokens = 0
+        tpos_sign_mul = -1 if track_in_reverse else 1
+        # Step 1: condition the visual features of the current frame on previous memories
+        if not is_init_cond_frame:
+            # Retrieve the memories encoded with the maskmem backbone
+            to_cat_memory, to_cat_memory_pos_embed = [], []
+            # Add conditioning frames's output first (all cond frames have t_pos=0 for
+            # when getting temporal positional embedding below)
+            assert len(output_dict["cond_frame_outputs"]) > 0
+            # Select a maximum number of temporally closest cond frames for cross attention
+            cond_outputs = output_dict["cond_frame_outputs"]
+            selected_cond_outputs, unselected_cond_outputs = select_closest_cond_frames(
+                frame_idx, cond_outputs, self.max_cond_frames_in_attn)
+            t_pos_and_prevs = [(0, out) for out in selected_cond_outputs.values()]
+            # Add last (self.num_maskmem - 1) frames before current frame for non-conditioning memory
+            # the earliest one has t_pos=1 and the latest one has t_pos=self.num_maskmem-1
+            # We also allow taking the memory frame non-consecutively (with stride>1), in which case
+            # we take (self.num_maskmem - 2) frames among every stride-th frames plus the last frame.
+            stride = 1 if self.training else self.memory_temporal_stride_for_eval
+            for t_pos in range(1, self.num_maskmem):
+                t_rel = self.num_maskmem - t_pos  # how many frames before current frame
+                if t_rel == 1:
+                    # for t_rel == 1, we take the last frame (regardless of r)
+                    if not track_in_reverse:
+                        # the frame immediately before this frame (i.e. frame_idx - 1)
+                        prev_frame_idx = frame_idx - t_rel
+                    else:
+                        # the frame immediately after this frame (i.e. frame_idx + 1)
+                        prev_frame_idx = frame_idx + t_rel
+                else:
+                    # for t_rel >= 2, we take the memory frame from every r-th frames
+                    if not track_in_reverse:
+                        # first find the nearest frame among every r-th frames before this frame
+                        # for r=1, this would be (frame_idx - 2)
+                        prev_frame_idx = ((frame_idx - 2) // stride) * stride
+                        # then seek further among every r-th frames
+                        prev_frame_idx = prev_frame_idx - (t_rel - 2) * stride
+                    else:
+                        # first find the nearest frame among every r-th frames after this frame
+                        # for r=1, this would be (frame_idx + 2)
+                        prev_frame_idx = -(-(frame_idx + 2) // stride) * stride
+                        # then seek further among every r-th frames
+                        prev_frame_idx = prev_frame_idx + (t_rel - 2) * stride
+                out = output_dict["non_cond_frame_outputs"].get(prev_frame_idx, None)
+                if out is None:
+                    # If an unselected conditioning frame is among the last (self.num_maskmem - 1)
+                    # frames, we still attend to it as if it's a non-conditioning frame.
+                    out = unselected_cond_outputs.get(prev_frame_idx, None)
+                t_pos_and_prevs.append((t_pos, out))
+
+            for t_pos, prev in t_pos_and_prevs:
+                if prev is None:
+                    continue  # skip padding frames
+                # "maskmem_features" might have been offloaded to CPU in demo use cases,
+                # so we load it back to GPU (it's a no-op if it's already on GPU).
+                feats = prev["maskmem_features"].to(device, non_blocking=True)
+                to_cat_memory.append(feats.flatten(2).permute(2, 0, 1))
+                # Spatial positional encoding (it might have been offloaded to CPU in eval)
+                maskmem_enc = prev["maskmem_pos_enc"][-1].to(device)
+                maskmem_enc = maskmem_enc.flatten(2).permute(2, 0, 1)
+                # Temporal positional encoding
+                maskmem_enc = (maskmem_enc + self.maskmem_tpos_enc[self.num_maskmem - t_pos - 1])
+                to_cat_memory_pos_embed.append(maskmem_enc)
+
+            # Construct the list of past object pointers
+            if self.use_obj_ptrs_in_encoder:
+                max_obj_ptrs_in_encoder = min(num_frames, self.max_obj_ptrs_in_encoder)
+                # First add those object pointers from selected conditioning frames
+                # (optionally, only include object pointers in the past during evaluation)
+                if not self.training and self.only_obj_ptrs_in_the_past_for_eval:
+                    ptr_cond_outputs = {
+                        t: out
+                        for t, out in selected_cond_outputs.items()
+                        if (t >= frame_idx if track_in_reverse else t <= frame_idx)
+                    }
+                else:
+                    ptr_cond_outputs = selected_cond_outputs
+                pos_and_ptrs = [
+                    # Temporal pos encoding contains how far away each pointer is from current frame
+                    (
+                        ((frame_idx - t) * tpos_sign_mul if self.use_signed_tpos_enc_to_obj_ptrs else abs(frame_idx -
+                                                                                                          t)),
+                        out["obj_ptr"],
+                    ) for t, out in ptr_cond_outputs.items()
+                ]
+                # Add up to (max_obj_ptrs_in_encoder - 1) non-conditioning frames before current frame
+                for t_diff in range(1, max_obj_ptrs_in_encoder):
+                    t = frame_idx + t_diff if track_in_reverse else frame_idx - t_diff
+                    if t < 0 or (num_frames is not None and t >= num_frames):
+                        break
+                    out = output_dict["non_cond_frame_outputs"].get(t, unselected_cond_outputs.get(t, None))
+                    if out is not None:
+                        pos_and_ptrs.append((t_diff, out["obj_ptr"]))
+                # If we have at least one object pointer, add them to the across attention
+                if len(pos_and_ptrs) > 0:
+                    pos_list, ptrs_list = zip(*pos_and_ptrs)
+                    # stack object pointers along dim=0 into [ptr_seq_len, B, C] shape
+                    obj_ptrs = torch.stack(ptrs_list, dim=0)
+                    # a temporal positional embedding based on how far each object pointer is from
+                    # the current frame (sine embedding normalized by the max pointer num).
+                    if self.add_tpos_enc_to_obj_ptrs:
+                        t_diff_max = max_obj_ptrs_in_encoder - 1
+                        tpos_dim = C if self.proj_tpos_enc_in_obj_ptrs else self.mem_dim
+                        obj_pos = torch.tensor(pos_list).to(device=device, non_blocking=True)
+                        obj_pos = get_1d_sine_pe(obj_pos / t_diff_max, dim=tpos_dim)
+                        obj_pos = self.obj_ptr_tpos_proj(obj_pos.to(self.obj_ptr_tpos_proj.weight.dtype))
+                        obj_pos = obj_pos.unsqueeze(1).expand(-1, B, self.mem_dim)
+                    else:
+                        obj_pos = obj_ptrs.new_zeros(len(pos_list), B, self.mem_dim)
+                    if self.mem_dim < C:
+                        # split a pointer into (C // self.mem_dim) tokens for self.mem_dim < C
+                        obj_ptrs = obj_ptrs.reshape(-1, B, C // self.mem_dim, self.mem_dim)
+                        obj_ptrs = obj_ptrs.permute(0, 2, 1, 3).flatten(0, 1)
+                        obj_pos = obj_pos.repeat_interleave(C // self.mem_dim, dim=0)
+                    to_cat_memory.append(obj_ptrs)
+                    to_cat_memory_pos_embed.append(obj_pos)
+                    num_obj_ptr_tokens = obj_ptrs.shape[0]
+                else:
+                    num_obj_ptr_tokens = 0
+        else:
+            # for initial conditioning frames, encode them without using any previous memory
+            if self.directly_add_no_mem_embed:
+                # directly add no-mem embedding (instead of using the transformer encoder)
+                pix_feat_with_mem = current_vision_feats[-1] + self.no_mem_embed
+                pix_feat_with_mem = pix_feat_with_mem.permute(1, 2, 0).view(B, C, H, W)
+                return pix_feat_with_mem
+
+            # Use a dummy token on the first frame (to avoid empty memory input to tranformer encoder)
+            to_cat_memory = [self.no_mem_embed.expand(1, B, self.mem_dim)]
+            to_cat_memory_pos_embed = [self.no_mem_pos_enc.expand(1, B, self.mem_dim)]
+
+        # Step 2: Concatenate the memories and forward through the transformer encoder
+        memory = torch.cat(to_cat_memory, dim=0)
+        memory_pos_embed = torch.cat(to_cat_memory_pos_embed, dim=0)
+
+        pix_feat_with_mem = self.memory_attention(
+            curr=current_vision_feats,
+            curr_pos=current_vision_pos_embeds,
+            memory=memory,
+            memory_pos=memory_pos_embed,
+            num_obj_ptr_tokens=num_obj_ptr_tokens,
+        )
+        # reshape the output (HW)BC => BCHW
+        pix_feat_with_mem = pix_feat_with_mem.permute(1, 2, 0).view(B, C, H, W)
+        return pix_feat_with_mem
+
+    def _encode_new_memory(
+        self,
+        current_vision_feats,
+        feat_sizes,
+        pred_masks_high_res,
+        object_score_logits,
+        is_mask_from_pts,
+    ):
+        """Encode the current image and its prediction into a memory feature."""
+        B = current_vision_feats[-1].size(1)  # batch size on this frame
+        C = self.hidden_dim
+        H, W = feat_sizes[-1]  # top-level (lowest-resolution) feature size
+        # top-level feature, (HW)BC => BCHW
+        pix_feat = current_vision_feats[-1].permute(1, 2, 0).view(B, C, H, W)
+        if self.non_overlap_masks_for_mem_enc and not self.training:
+            # optionally, apply non-overlapping constraints to the masks (it's applied
+            # in the batch dimension and should only be used during eval, where all
+            # the objects come from the same video under batch size 1).
+            pred_masks_high_res = self._apply_non_overlapping_constraints(pred_masks_high_res)
+        # scale the raw mask logits with a temperature before applying sigmoid
+        binarize = self.binarize_mask_from_pts_for_mem_enc and is_mask_from_pts
+        if binarize and not self.training:
+            mask_for_mem = (pred_masks_high_res > 0).to(pred_masks_high_res.dtype)
+        else:
+            # apply sigmoid on the raw mask logits to turn them into range (0, 1)
+            mask_for_mem = torch.sigmoid(pred_masks_high_res)
+        # apply scale and bias terms to the sigmoid probabilities
+        if self.sigmoid_scale_for_mem_enc != 1.0:
+            mask_for_mem = mask_for_mem * self.sigmoid_scale_for_mem_enc
+        if self.sigmoid_bias_for_mem_enc != 0.0:
+            mask_for_mem = mask_for_mem + self.sigmoid_bias_for_mem_enc
+        maskmem_out = self.memory_encoder(
+            pix_feat,
+            mask_for_mem,
+            skip_mask_sigmoid=True  # sigmoid already applied
+        )
+        maskmem_features = maskmem_out["vision_features"]
+        maskmem_pos_enc = maskmem_out["vision_pos_enc"]
+        # add a no-object embedding to the spatial memory to indicate that the frame
+        # is predicted to be occluded (i.e. no object is appearing in the frame)
+        if self.no_obj_embed_spatial is not None:
+            is_obj_appearing = (object_score_logits > 0).to(object_score_logits.dtype)
+            maskmem_features += (1 - is_obj_appearing[..., None, None]
+                                 ) * self.no_obj_embed_spatial[..., None, None].expand(*maskmem_features.shape)
+
+        return maskmem_features, maskmem_pos_enc
+
+    def _track_step(
+        self,
+        frame_idx,
+        is_init_cond_frame,
+        current_vision_feats,
+        current_vision_pos_embeds,
+        feat_sizes,
+        point_inputs,
+        mask_inputs,
+        hidden_inputs,
+        output_dict,
+        num_frames,
+        track_in_reverse,
+        prev_sam_mask_logits,
+    ):
+        current_out = {"point_inputs": point_inputs, "mask_inputs": mask_inputs, "hidden_inputs": hidden_inputs}
+        # High-resolution feature maps for the SAM head, reshape (HW)BC => BCHW
+        if len(current_vision_feats) > 1:
+            high_res_features = [
+                x.permute(1, 2, 0).view(x.size(1), x.size(2), *s)
+                for x, s in zip(current_vision_feats[:-1], feat_sizes[:-1])
+            ]
+        else:
+            high_res_features = None
+        if mask_inputs is not None and self.use_mask_input_as_output_without_sam:
+            # When use_mask_input_as_output_without_sam=True, we directly output the mask input
+            # (see it as a GT mask) without using a SAM prompt encoder + mask decoder.
+            pix_feat = current_vision_feats[-1].permute(1, 2, 0)
+            pix_feat = pix_feat.view(-1, self.hidden_dim, *feat_sizes[-1])
+            sam_outputs = self._use_mask_as_output(pix_feat, high_res_features, mask_inputs)
+        else:
+            # fused the visual feature with previous memory features in the memory bank
+            pix_feat = self._prepare_memory_conditioned_features(
+                frame_idx=frame_idx,
+                is_init_cond_frame=is_init_cond_frame,
+                current_vision_feats=current_vision_feats[-1:],
+                current_vision_pos_embeds=current_vision_pos_embeds[-1:],
+                feat_sizes=feat_sizes[-1:],
+                output_dict=output_dict,
+                num_frames=num_frames,
+                track_in_reverse=track_in_reverse,
+            )
+            # apply SAM-style segmentation head
+            # here we might feed previously predicted low-res SAM mask logits into the SAM mask decoder,
+            # e.g. in demo where such logits come from earlier interaction instead of correction sampling
+            # (in this case, any `mask_inputs` shouldn't reach here as they are sent to _use_mask_as_output instead)
+            if prev_sam_mask_logits is not None:
+                assert point_inputs is not None and mask_inputs is None
+                mask_inputs = prev_sam_mask_logits
+            multimask_output = self._use_multimask(is_init_cond_frame, point_inputs)
+            sam_outputs = self._forward_sam_heads(
+                backbone_features=pix_feat,
+                point_inputs=point_inputs,
+                mask_inputs=mask_inputs,
+                hidden_inputs=hidden_inputs,
+                high_res_features=high_res_features,
+                multimask_output=multimask_output,
+            )
+
+        return current_out, sam_outputs, high_res_features, pix_feat
+
+    def _encode_memory_in_output(
+        self,
+        current_vision_feats,
+        feat_sizes,
+        point_inputs,
+        run_mem_encoder,
+        high_res_masks,
+        object_score_logits,
+        current_out,
+    ):
+        if run_mem_encoder and self.num_maskmem > 0:
+            high_res_masks_for_mem_enc = high_res_masks
+            maskmem_features, maskmem_pos_enc = self._encode_new_memory(
+                current_vision_feats=current_vision_feats,
+                feat_sizes=feat_sizes,
+                pred_masks_high_res=high_res_masks_for_mem_enc,
+                object_score_logits=object_score_logits,
+                is_mask_from_pts=(point_inputs is not None),
+            )
+            current_out["maskmem_features"] = maskmem_features
+            current_out["maskmem_pos_enc"] = maskmem_pos_enc
+        else:
+            current_out["maskmem_features"] = None
+            current_out["maskmem_pos_enc"] = None
+
+    def track_step(
+        self,
+        frame_idx,
+        is_init_cond_frame,
+        current_vision_feats,
+        current_vision_pos_embeds,
+        feat_sizes,
+        point_inputs,
+        mask_inputs,
+        hidden_inputs,
+        output_dict,
+        num_frames,
+        track_in_reverse=False,  # tracking in reverse time order (for demo usage)
+        # Whether to run the memory encoder on the predicted masks. Sometimes we might want
+        # to skip the memory encoder with `run_mem_encoder=False`. For example,
+        # in demo we might call `track_step` multiple times for each user click,
+        # and only encode the memory when the user finalizes their clicks. And in ablation
+        # settings like SAM training on static images, we don't need the memory encoder.
+        run_mem_encoder=True,
+        # The previously predicted SAM mask logits (which can be fed together with new clicks in demo).
+        prev_sam_mask_logits=None,
+    ):
+        current_out, sam_outputs, _, _ = self._track_step(
+            frame_idx,
+            is_init_cond_frame,
+            current_vision_feats,
+            current_vision_pos_embeds,
+            feat_sizes,
+            point_inputs,
+            mask_inputs,
+            hidden_inputs,
+            output_dict,
+            num_frames,
+            track_in_reverse,
+            prev_sam_mask_logits,
+        )
+
+        (
+            _,
+            _,
+            _,
+            low_res_masks,
+            high_res_masks,
+            obj_ptr,
+            object_score_logits,
+        ) = sam_outputs
+
+        current_out["pred_masks"] = low_res_masks
+        current_out["pred_masks_high_res"] = high_res_masks
+        current_out["obj_ptr"] = obj_ptr
+        if not self.training:
+            # Only add this in inference (to avoid unused param in activation checkpointing;
+            # it's mainly used in the demo to encode spatial memories w/ consolidated masks)
+            current_out["object_score_logits"] = object_score_logits
+
+        # Finally run the memory encoder on the predicted mask to encode
+        # it into a new memory feature (that can be used in future frames)
+        self._encode_memory_in_output(
+            current_vision_feats,
+            feat_sizes,
+            point_inputs,
+            run_mem_encoder,
+            high_res_masks,
+            object_score_logits,
+            current_out,
+        )
+
+        return current_out
+
+    def _use_multimask(self, is_init_cond_frame, point_inputs):
+        """Whether to use multimask output in the SAM head."""
+        num_pts = 0 if point_inputs is None else point_inputs["point_labels"].size(1)
+        multimask_output = (
+            self.multimask_output_in_sam and (is_init_cond_frame or self.multimask_output_for_tracking)
+            and (self.multimask_min_pt_num <= num_pts <= self.multimask_max_pt_num))
+        return multimask_output
+
+    def _apply_non_overlapping_constraints(self, pred_masks):
+        """
+        Apply non-overlapping constraints to the object scores in pred_masks. Here we
+        keep only the highest scoring object at each spatial location in pred_masks.
+        """
+        batch_size = pred_masks.size(0)
+        if batch_size == 1:
+            return pred_masks
+
+        device = pred_masks.device
+        # "max_obj_inds": object index of the object with the highest score at each location
+        max_obj_inds = torch.argmax(pred_masks, dim=0, keepdim=True)
+        # "batch_obj_inds": object index of each object slice (along dim 0) in `pred_masks`
+        batch_obj_inds = torch.arange(batch_size, device=device)[:, None, None, None]
+        keep = max_obj_inds == batch_obj_inds
+        # suppress overlapping regions' scores below -10.0 so that the foreground regions
+        # don't overlap (here sigmoid(-10.0)=4.5398e-05)
+        pred_masks = torch.where(keep, pred_masks, torch.clamp(pred_masks, max=-10.0))
+        return pred_masks

sam2/modeling/sam2_utils.py

@@ -0,0 +1,320 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.
+
+import copy
+from typing import Tuple
+
+import numpy as np
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+
+from sam2.utils.misc import mask_to_box
+
+
+def select_closest_cond_frames(frame_idx, cond_frame_outputs, max_cond_frame_num):
+    """
+    Select up to `max_cond_frame_num` conditioning frames from `cond_frame_outputs`
+    that are temporally closest to the current frame at `frame_idx`. Here, we take
+    - a) the closest conditioning frame before `frame_idx` (if any);
+    - b) the closest conditioning frame after `frame_idx` (if any);
+    - c) any other temporally closest conditioning frames until reaching a total
+         of `max_cond_frame_num` conditioning frames.
+
+    Outputs:
+    - selected_outputs: selected items (keys & values) from `cond_frame_outputs`.
+    - unselected_outputs: items (keys & values) not selected in `cond_frame_outputs`.
+    """
+    if max_cond_frame_num == -1 or len(cond_frame_outputs) <= max_cond_frame_num:
+        selected_outputs = cond_frame_outputs
+        unselected_outputs = {}
+    else:
+        assert max_cond_frame_num >= 2, "we should allow using 2+ conditioning frames"
+        selected_outputs = {}
+
+        # the closest conditioning frame before `frame_idx` (if any)
+        idx_before = max((t for t in cond_frame_outputs if t < frame_idx), default=None)
+        if idx_before is not None:
+            selected_outputs[idx_before] = cond_frame_outputs[idx_before]
+
+        # the closest conditioning frame after `frame_idx` (if any)
+        idx_after = min((t for t in cond_frame_outputs if t >= frame_idx), default=None)
+        if idx_after is not None:
+            selected_outputs[idx_after] = cond_frame_outputs[idx_after]
+
+        # add other temporally closest conditioning frames until reaching a total
+        # of `max_cond_frame_num` conditioning frames.
+        num_remain = max_cond_frame_num - len(selected_outputs)
+        inds_remain = sorted(
+            (t for t in cond_frame_outputs if t not in selected_outputs),
+            key=lambda x: abs(x - frame_idx),
+        )[:num_remain]
+        selected_outputs.update((t, cond_frame_outputs[t]) for t in inds_remain)
+        unselected_outputs = {t: v for t, v in cond_frame_outputs.items() if t not in selected_outputs}
+
+    return selected_outputs, unselected_outputs
+
+
+def get_1d_sine_pe(pos_inds, dim, temperature=10000):
+    """
+    Get 1D sine positional embedding as in the original Transformer paper.
+    """
+    pe_dim = dim // 2
+    dim_t = torch.arange(pe_dim, dtype=torch.float32, device=pos_inds.device)
+    dim_t = temperature**(2 * (dim_t // 2) / pe_dim)
+
+    pos_embed = pos_inds.unsqueeze(-1) / dim_t
+    pos_embed = torch.cat([pos_embed.sin(), pos_embed.cos()], dim=-1)
+    return pos_embed
+
+
+def get_activation_fn(activation):
+    """Return an activation function given a string"""
+    if activation == "relu":
+        return F.relu
+    if activation == "gelu":
+        return F.gelu
+    if activation == "glu":
+        return F.glu
+    raise RuntimeError(f"activation should be relu/gelu, not {activation}.")
+
+
+def get_clones(module, N):
+    return nn.ModuleList([copy.deepcopy(module) for i in range(N)])
+
+
+class DropPath(nn.Module):
+    # adapted from https://github.com/huggingface/pytorch-image-models/blob/main/timm/layers/drop.py
+    def __init__(self, drop_prob=0.0, scale_by_keep=True):
+        super(DropPath, self).__init__()
+        self.drop_prob = drop_prob
+        self.scale_by_keep = scale_by_keep
+
+    def forward(self, x):
+        if self.drop_prob == 0.0 or not self.training:
+            return x
+        keep_prob = 1 - self.drop_prob
+        shape = (x.shape[0], ) + (1, ) * (x.ndim - 1)
+        random_tensor = x.new_empty(shape).bernoulli_(keep_prob)
+        if keep_prob > 0.0 and self.scale_by_keep:
+            random_tensor.div_(keep_prob)
+        return x * random_tensor
+
+
+# Lightly adapted from
+# https://github.com/facebookresearch/MaskFormer/blob/main/mask_former/modeling/transformer/transformer_predictor.py # noqa
+class MLP(nn.Module):
+
+    def __init__(
+        self,
+        input_dim: int,
+        hidden_dim: int,
+        output_dim: int,
+        num_layers: int,
+        activation: nn.Module = nn.ReLU,
+        sigmoid_output: bool = False,
+    ) -> None:
+        super().__init__()
+        self.num_layers = num_layers
+        h = [hidden_dim] * (num_layers - 1)
+        self.layers = nn.ModuleList(nn.Linear(n, k) for n, k in zip([input_dim] + h, h + [output_dim]))
+        self.sigmoid_output = sigmoid_output
+        self.act = activation()
+
+    def forward(self, x):
+        for i, layer in enumerate(self.layers):
+            x = self.act(layer(x)) if i < self.num_layers - 1 else layer(x)
+        if self.sigmoid_output:
+            x = F.sigmoid(x)
+        return x
+
+
+# From https://github.com/facebookresearch/detectron2/blob/main/detectron2/layers/batch_norm.py # noqa
+# Itself from https://github.com/facebookresearch/ConvNeXt/blob/d1fa8f6fef0a165b27399986cc2bdacc92777e40/models/convnext.py#L119  # noqa
+class LayerNorm2d(nn.Module):
+
+    def __init__(self, num_channels: int, eps: float = 1e-6) -> None:
+        super().__init__()
+        self.weight = nn.Parameter(torch.ones(num_channels))
+        self.bias = nn.Parameter(torch.zeros(num_channels))
+        self.eps = eps
+
+    def forward(self, x: torch.Tensor) -> torch.Tensor:
+        u = x.mean(1, keepdim=True)
+        s = (x - u).pow(2).mean(1, keepdim=True)
+        x = (x - u) / torch.sqrt(s + self.eps)
+        x = self.weight[:, None, None] * x + self.bias[:, None, None]
+        return x
+
+
+def sample_box_points(
+    masks: torch.Tensor,
+    noise: float = 0.1,  # SAM default
+    noise_bound: int = 20,  # SAM default
+    top_left_label: int = 2,
+    bottom_right_label: int = 3,
+) -> Tuple[np.array, np.array]:
+    """
+    Sample a noised version of the top left and bottom right corners of a given `bbox`
+
+    Inputs:
+    - masks: [B, 1, H,W] boxes, dtype=torch.Tensor
+    - noise: noise as a fraction of box width and height, dtype=float
+    - noise_bound: maximum amount of noise (in pure pixesl), dtype=int
+
+    Returns:
+    - box_coords: [B, num_pt, 2], contains (x, y) coordinates of top left and bottom right box corners, dtype=torch.float
+    - box_labels: [B, num_pt], label 2 is reserverd for top left and 3 for bottom right corners, dtype=torch.int32
+    """
+    device = masks.device
+    box_coords = mask_to_box(masks)
+    B, _, H, W = masks.shape
+    box_labels = torch.tensor([top_left_label, bottom_right_label], dtype=torch.int, device=device).repeat(B)
+    if noise > 0.0:
+        if not isinstance(noise_bound, torch.Tensor):
+            noise_bound = torch.tensor(noise_bound, device=device)
+        bbox_w = box_coords[..., 2] - box_coords[..., 0]
+        bbox_h = box_coords[..., 3] - box_coords[..., 1]
+        max_dx = torch.min(bbox_w * noise, noise_bound)
+        max_dy = torch.min(bbox_h * noise, noise_bound)
+        box_noise = 2 * torch.rand(B, 1, 4, device=device) - 1
+        box_noise = box_noise * torch.stack((max_dx, max_dy, max_dx, max_dy), dim=-1)
+
+        box_coords = box_coords + box_noise
+        img_bounds = (torch.tensor([W, H, W, H], device=device) - 1)  # uncentered pixel coords
+        box_coords.clamp_(torch.zeros_like(img_bounds), img_bounds)  # In place clamping
+
+    box_coords = box_coords.reshape(-1, 2, 2)  # always 2 points
+    box_labels = box_labels.reshape(-1, 2)
+    return box_coords, box_labels
+
+
+def sample_random_points_from_errors(gt_masks, pred_masks, num_pt=1, positive_only=False):
+    """
+    Sample `num_pt` random points (along with their labels) independently from the error regions.
+
+    Inputs:
+    - gt_masks: [B, 1, H_im, W_im] masks, dtype=torch.bool
+    - pred_masks: [B, 1, H_im, W_im] masks, dtype=torch.bool or None
+    - num_pt: int, number of points to sample independently for each of the B error maps
+
+    Outputs:
+    - points: [B, num_pt, 2], dtype=torch.float, contains (x, y) coordinates of each sampled point
+    - labels: [B, num_pt], dtype=torch.int32, where 1 means positive clicks and 0 means
+      negative clicks
+    """
+    if pred_masks is None:  # if pred_masks is not provided, treat it as empty
+        pred_masks = torch.zeros_like(gt_masks)
+    assert gt_masks.dtype == torch.bool and gt_masks.size(1) == 1
+    assert pred_masks.dtype == torch.bool and pred_masks.shape == gt_masks.shape
+    assert num_pt >= 0
+
+    B, _, H_im, W_im = gt_masks.shape
+    device = gt_masks.device
+
+    # false positive region, a new point sampled in this region should have
+    # negative label to correct the FP error
+    fp_masks = ~gt_masks & pred_masks
+    # false negative region, a new point sampled in this region should have
+    # positive label to correct the FN error
+    fn_masks = gt_masks & ~pred_masks
+    # whether the prediction completely match the ground-truth on each mask
+    all_correct = torch.all((gt_masks == pred_masks).flatten(2), dim=2)
+    all_correct = all_correct[..., None, None]
+
+    # channel 0 is FP map, while channel 1 is FN map
+    pts_noise = torch.rand(B, num_pt, H_im, W_im, 2, device=device)
+    # sample a negative new click from FP region or a positive new click
+    # from FN region, depend on where the maximum falls,
+    # and in case the predictions are all correct (no FP or FN), we just
+    # sample a negative click from the background region
+    pts_noise[..., 0] *= fp_masks | (all_correct & ~gt_masks)
+    if positive_only:
+        pts_noise[..., 0] = -1
+    pts_noise[..., 1] *= fn_masks
+    pts_idx = pts_noise.flatten(2).argmax(dim=2)
+    labels = (pts_idx % 2).to(torch.int32)
+    pts_idx = pts_idx // 2
+    pts_x = pts_idx % W_im
+    pts_y = pts_idx // W_im
+    points = torch.stack([pts_x, pts_y], dim=2).to(torch.float)
+    return points, labels
+
+
+def sample_one_point_from_error_center(gt_masks, pred_masks, padding=True, positive_only=False):
+    """
+    Sample 1 random point (along with its label) from the center of each error region,
+    that is, the point with the largest distance to the boundary of each error region.
+    This is the RITM sampling method from https://github.com/saic-vul/ritm_interactive_segmentation/blob/master/isegm/inference/clicker.py
+
+    Inputs:
+    - gt_masks: [B, 1, H_im, W_im] masks, dtype=torch.bool
+    - pred_masks: [B, 1, H_im, W_im] masks, dtype=torch.bool or None
+    - padding: if True, pad with boundary of 1 px for distance transform
+
+    Outputs:
+    - points: [B, 1, 2], dtype=torch.float, contains (x, y) coordinates of each sampled point
+    - labels: [B, 1], dtype=torch.int32, where 1 means positive clicks and 0 means negative clicks
+    """
+    import cv2
+
+    if pred_masks is None:
+        pred_masks = torch.zeros_like(gt_masks)
+    assert gt_masks.dtype == torch.bool and gt_masks.size(1) == 1
+    assert pred_masks.dtype == torch.bool and pred_masks.shape == gt_masks.shape
+
+    B, _, _, W_im = gt_masks.shape
+    device = gt_masks.device
+
+    # false positive region, a new point sampled in this region should have
+    # negative label to correct the FP error
+    fp_masks = ~gt_masks & pred_masks
+    # false negative region, a new point sampled in this region should have
+    # positive label to correct the FN error
+    fn_masks = gt_masks & ~pred_masks
+
+    fp_masks = fp_masks.cpu().numpy()
+    fn_masks = fn_masks.cpu().numpy()
+    points = torch.zeros(B, 1, 2, dtype=torch.float)
+    labels = torch.ones(B, 1, dtype=torch.int32)
+    for b in range(B):
+        fn_mask = fn_masks[b, 0]
+        fp_mask = fp_masks[b, 0]
+        if padding:
+            fn_mask = np.pad(fn_mask, ((1, 1), (1, 1)), "constant")
+            fp_mask = np.pad(fp_mask, ((1, 1), (1, 1)), "constant")
+        # compute the distance of each point in FN/FP region to its boundary
+        fn_mask_dt = cv2.distanceTransform(fn_mask.astype(np.uint8), cv2.DIST_L2, 0)
+        fp_mask_dt = cv2.distanceTransform(fp_mask.astype(np.uint8), cv2.DIST_L2, 0)
+        if padding:
+            fn_mask_dt = fn_mask_dt[1:-1, 1:-1]
+            fp_mask_dt = fp_mask_dt[1:-1, 1:-1]
+
+        # take the point in FN/FP region with the largest distance to its boundary
+        fn_mask_dt_flat = fn_mask_dt.reshape(-1)
+        fp_mask_dt_flat = fp_mask_dt.reshape(-1)
+        fn_argmax = np.argmax(fn_mask_dt_flat)
+        fp_argmax = np.argmax(fp_mask_dt_flat)
+        is_positive = fn_mask_dt_flat[fn_argmax] > fp_mask_dt_flat[fp_argmax]
+        if positive_only:
+            is_positive = True
+        pt_idx = fn_argmax if is_positive else fp_argmax
+        points[b, 0, 0] = pt_idx % W_im  # x
+        points[b, 0, 1] = pt_idx // W_im  # y
+        labels[b, 0] = int(is_positive)
+
+    points = points.to(device)
+    labels = labels.to(device)
+    return points, labels
+
+
+def get_next_point(gt_masks, pred_masks, method, positive_only=True):
+    if method == "uniform":
+        return sample_random_points_from_errors(gt_masks, pred_masks, positive_only=positive_only)
+    elif method == "center":
+        return sample_one_point_from_error_center(gt_masks, pred_masks, positive_only=positive_only)
+    else:
+        raise ValueError(f"unknown sampling method {method}")

sam2/sam2_image_predictor.py

@@ -0,0 +1,428 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.
+
+import logging
+
+from typing import List, Optional, Tuple, Union
+
+import numpy as np
+import torch
+from PIL.Image import Image
+
+from sam2.modeling.sam2_base import SAM2Base
+
+from sam2.utils.transforms import SAM2Transforms
+
+
+class SAM2ImagePredictor:
+
+    def __init__(
+        self,
+        sam_model: SAM2Base,
+        mask_threshold=0.0,
+        max_hole_area=0.0,
+        max_sprinkle_area=0.0,
+        **kwargs,
+    ) -> None:
+        """
+        Uses SAM-2 to calculate the image embedding for an image, and then
+        allow repeated, efficient mask prediction given prompts.
+
+        Arguments:
+          sam_model (Sam-2): The model to use for mask prediction.
+          mask_threshold (float): The threshold to use when converting mask logits
+            to binary masks. Masks are thresholded at 0 by default.
+          max_hole_area (int): If max_hole_area > 0, we fill small holes in up to
+            the maximum area of max_hole_area in low_res_masks.
+          max_sprinkle_area (int): If max_sprinkle_area > 0, we remove small sprinkles up to
+            the maximum area of max_sprinkle_area in low_res_masks.
+        """
+        super().__init__()
+        self.model = sam_model
+        self._transforms = SAM2Transforms(
+            resolution=self.model.image_size,
+            mask_threshold=mask_threshold,
+            max_hole_area=max_hole_area,
+            max_sprinkle_area=max_sprinkle_area,
+        )
+
+        # Predictor state
+        self._is_image_set = False
+        self._features = None
+        self._orig_hw = None
+        # Whether the predictor is set for single image or a batch of images
+        self._is_batch = False
+
+        # Predictor config
+        self.mask_threshold = mask_threshold
+
+        # Spatial dim for backbone feature maps
+        self._bb_feat_sizes = [
+            (256, 256),
+            (128, 128),
+            (64, 64),
+        ]
+
+    @classmethod
+    def from_pretrained(cls, model_id: str, **kwargs) -> "SAM2ImagePredictor":
+        """
+        Load a pretrained model from the Hugging Face hub.
+
+        Arguments:
+          model_id (str): The Hugging Face repository ID.
+          **kwargs: Additional arguments to pass to the model constructor.
+
+        Returns:
+          (SAM2ImagePredictor): The loaded model.
+        """
+        from sam2.build_sam import build_sam2_hf
+
+        sam_model = build_sam2_hf(model_id, **kwargs)
+        return cls(sam_model, **kwargs)
+
+    @torch.no_grad()
+    def set_image(
+        self,
+        image: Union[np.ndarray, Image],
+    ) -> None:
+        """
+        Calculates the image embeddings for the provided image, allowing
+        masks to be predicted with the 'predict' method.
+
+        Arguments:
+          image (np.ndarray or PIL Image): The input image to embed in RGB format. The image should be in HWC format if np.ndarray, or WHC format if PIL Image
+          with pixel values in [0, 255].
+          image_format (str): The color format of the image, in ['RGB', 'BGR'].
+        """
+        self.reset_predictor()
+        # Transform the image to the form expected by the model
+        if isinstance(image, np.ndarray):
+            logging.info("For numpy array image, we assume (HxWxC) format")
+            self._orig_hw = [image.shape[:2]]
+        elif isinstance(image, Image):
+            w, h = image.size
+            self._orig_hw = [(h, w)]
+        else:
+            raise NotImplementedError("Image format not supported")
+
+        input_image = self._transforms(image)
+        input_image = input_image[None, ...].to(self.device)
+
+        assert (len(input_image.shape) == 4
+                and input_image.shape[1] == 3), f"input_image must be of size 1x3xHxW, got {input_image.shape}"
+        logging.info("Computing image embeddings for the provided image...")
+        backbone_out = self.model.forward_image(input_image)
+        _, vision_feats, _, _ = self.model._prepare_backbone_features(backbone_out)
+        # Add no_mem_embed, which is added to the lowest rest feat. map during training on videos
+        if self.model.directly_add_no_mem_embed:
+            vision_feats[-1] = vision_feats[-1] + self.model.no_mem_embed
+
+        feats = [
+            feat.permute(1, 2, 0).view(1, -1, *feat_size)
+            for feat, feat_size in zip(vision_feats[::-1], self._bb_feat_sizes[::-1])
+        ][::-1]
+        self._features = {"image_embed": feats[-1], "high_res_feats": feats[:-1]}
+        self._is_image_set = True
+        logging.info("Image embeddings computed.")
+
+    @torch.no_grad()
+    def set_image_batch(
+        self,
+        image_list: List[Union[np.ndarray]],
+    ) -> None:
+        """
+        Calculates the image embeddings for the provided image batch, allowing
+        masks to be predicted with the 'predict_batch' method.
+
+        Arguments:
+          image_list (List[np.ndarray]): The input images to embed in RGB format. The image should be in HWC format if np.ndarray
+          with pixel values in [0, 255].
+        """
+        self.reset_predictor()
+        assert isinstance(image_list, list)
+        self._orig_hw = []
+        for image in image_list:
+            assert isinstance(image,
+                              np.ndarray), "Images are expected to be an np.ndarray in RGB format, and of shape  HWC"
+            self._orig_hw.append(image.shape[:2])
+        # Transform the image to the form expected by the model
+        img_batch = self._transforms.forward_batch(image_list)
+        img_batch = img_batch.to(self.device)
+        batch_size = img_batch.shape[0]
+        assert (len(img_batch.shape) == 4
+                and img_batch.shape[1] == 3), f"img_batch must be of size Bx3xHxW, got {img_batch.shape}"
+        logging.info("Computing image embeddings for the provided images...")
+        backbone_out = self.model.forward_image(img_batch)
+        _, vision_feats, _, _ = self.model._prepare_backbone_features(backbone_out)
+        # Add no_mem_embed, which is added to the lowest rest feat. map during training on videos
+        if self.model.directly_add_no_mem_embed:
+            vision_feats[-1] = vision_feats[-1] + self.model.no_mem_embed
+
+        feats = [
+            feat.permute(1, 2, 0).view(batch_size, -1, *feat_size)
+            for feat, feat_size in zip(vision_feats[::-1], self._bb_feat_sizes[::-1])
+        ][::-1]
+        self._features = {"image_embed": feats[-1], "high_res_feats": feats[:-1]}
+        self._is_image_set = True
+        self._is_batch = True
+        logging.info("Image embeddings computed.")
+
+    def predict_batch(
+        self,
+        point_coords_batch: List[np.ndarray] = None,
+        point_labels_batch: List[np.ndarray] = None,
+        box_batch: List[np.ndarray] = None,
+        mask_input_batch: List[np.ndarray] = None,
+        multimask_output: bool = True,
+        return_logits: bool = False,
+        normalize_coords=True,
+    ) -> Tuple[List[np.ndarray], List[np.ndarray], List[np.ndarray]]:
+        """This function is very similar to predict(...), however it is used for batched mode, when the model is expected to generate predictions on multiple images.
+        It returns a tuple of lists of masks, ious, and low_res_masks_logits.
+        """
+        assert self._is_batch, "This function should only be used when in batched mode"
+        if not self._is_image_set:
+            raise RuntimeError("An image must be set with .set_image_batch(...) before mask prediction.")
+        num_images = len(self._features["image_embed"])
+        all_masks = []
+        all_ious = []
+        all_low_res_masks = []
+        for img_idx in range(num_images):
+            # Transform input prompts
+            point_coords = (point_coords_batch[img_idx] if point_coords_batch is not None else None)
+            point_labels = (point_labels_batch[img_idx] if point_labels_batch is not None else None)
+            box = box_batch[img_idx] if box_batch is not None else None
+            mask_input = (mask_input_batch[img_idx] if mask_input_batch is not None else None)
+            mask_input, unnorm_coords, labels, unnorm_box = self._prep_prompts(
+                point_coords,
+                point_labels,
+                box,
+                mask_input,
+                normalize_coords,
+                img_idx=img_idx,
+            )
+            masks, iou_predictions, low_res_masks = self._predict(
+                unnorm_coords,
+                labels,
+                unnorm_box,
+                mask_input,
+                multimask_output,
+                return_logits=return_logits,
+                img_idx=img_idx,
+            )
+            masks_np = masks.squeeze(0).float().detach().cpu().numpy()
+            iou_predictions_np = (iou_predictions.squeeze(0).float().detach().cpu().numpy())
+            low_res_masks_np = low_res_masks.squeeze(0).float().detach().cpu().numpy()
+            all_masks.append(masks_np)
+            all_ious.append(iou_predictions_np)
+            all_low_res_masks.append(low_res_masks_np)
+
+        return all_masks, all_ious, all_low_res_masks
+
+    def predict(
+        self,
+        point_coords: Optional[np.ndarray] = None,
+        point_labels: Optional[np.ndarray] = None,
+        box: Optional[np.ndarray] = None,
+        mask_input: Optional[np.ndarray] = None,
+        multimask_output: bool = True,
+        return_logits: bool = False,
+        normalize_coords=True,
+    ) -> Tuple[np.ndarray, np.ndarray, np.ndarray]:
+        """
+        Predict masks for the given input prompts, using the currently set image.
+
+        Arguments:
+          point_coords (np.ndarray or None): A Nx2 array of point prompts to the
+            model. Each point is in (X,Y) in pixels.
+          point_labels (np.ndarray or None): A length N array of labels for the
+            point prompts. 1 indicates a foreground point and 0 indicates a
+            background point.
+          box (np.ndarray or None): A length 4 array given a box prompt to the
+            model, in XYXY format.
+          mask_input (np.ndarray): A low resolution mask input to the model, typically
+            coming from a previous prediction iteration. Has form 1xHxW, where
+            for SAM, H=W=256.
+          multimask_output (bool): If true, the model will return three masks.
+            For ambiguous input prompts (such as a single click), this will often
+            produce better masks than a single prediction. If only a single
+            mask is needed, the model's predicted quality score can be used
+            to select the best mask. For non-ambiguous prompts, such as multiple
+            input prompts, multimask_output=False can give better results.
+          return_logits (bool): If true, returns un-thresholded masks logits
+            instead of a binary mask.
+          normalize_coords (bool): If true, the point coordinates will be normalized to the range [0,1] and point_coords is expected to be wrt. image dimensions.
+
+        Returns:
+          (np.ndarray): The output masks in CxHxW format, where C is the
+            number of masks, and (H, W) is the original image size.
+          (np.ndarray): An array of length C containing the model's
+            predictions for the quality of each mask.
+          (np.ndarray): An array of shape CxHxW, where C is the number
+            of masks and H=W=256. These low resolution logits can be passed to
+            a subsequent iteration as mask input.
+        """
+        if not self._is_image_set:
+            raise RuntimeError("An image must be set with .set_image(...) before mask prediction.")
+
+        # Transform input prompts
+
+        mask_input, unnorm_coords, labels, unnorm_box = self._prep_prompts(point_coords, point_labels, box, mask_input,
+                                                                           normalize_coords)
+
+        masks, iou_predictions, low_res_masks = self._predict(
+            unnorm_coords,
+            labels,
+            unnorm_box,
+            mask_input,
+            multimask_output,
+            return_logits=return_logits,
+        )
+
+        masks_np = masks.squeeze(0).float().detach().cpu().numpy()
+        iou_predictions_np = iou_predictions.squeeze(0).float().detach().cpu().numpy()
+        low_res_masks_np = low_res_masks.squeeze(0).float().detach().cpu().numpy()
+        return masks_np, iou_predictions_np, low_res_masks_np
+
+    def _prep_prompts(self, point_coords, point_labels, box, mask_logits, normalize_coords, img_idx=-1):
+
+        unnorm_coords, labels, unnorm_box, mask_input = None, None, None, None
+        if point_coords is not None:
+            assert (point_labels is not None), "point_labels must be supplied if point_coords is supplied."
+            point_coords = torch.as_tensor(point_coords, dtype=torch.float, device=self.device)
+            unnorm_coords = self._transforms.transform_coords(
+                point_coords, normalize=normalize_coords, orig_hw=self._orig_hw[img_idx])
+            labels = torch.as_tensor(point_labels, dtype=torch.int, device=self.device)
+            if len(unnorm_coords.shape) == 2:
+                unnorm_coords, labels = unnorm_coords[None, ...], labels[None, ...]
+        if box is not None:
+            box = torch.as_tensor(box, dtype=torch.float, device=self.device)
+            unnorm_box = self._transforms.transform_boxes(
+                box, normalize=normalize_coords, orig_hw=self._orig_hw[img_idx])  # Bx2x2
+        if mask_logits is not None:
+            mask_input = torch.as_tensor(mask_logits, dtype=torch.float, device=self.device)
+            if len(mask_input.shape) == 3:
+                mask_input = mask_input[None, :, :, :]
+        return mask_input, unnorm_coords, labels, unnorm_box
+
+    @torch.no_grad()
+    def _predict(
+        self,
+        point_coords: Optional[torch.Tensor],
+        point_labels: Optional[torch.Tensor],
+        boxes: Optional[torch.Tensor] = None,
+        mask_input: Optional[torch.Tensor] = None,
+        multimask_output: bool = True,
+        return_logits: bool = False,
+        img_idx: int = -1,
+    ) -> Tuple[torch.Tensor, torch.Tensor, torch.Tensor]:
+        """
+        Predict masks for the given input prompts, using the currently set image.
+        Input prompts are batched torch tensors and are expected to already be
+        transformed to the input frame using SAM2Transforms.
+
+        Arguments:
+          point_coords (torch.Tensor or None): A BxNx2 array of point prompts to the
+            model. Each point is in (X,Y) in pixels.
+          point_labels (torch.Tensor or None): A BxN array of labels for the
+            point prompts. 1 indicates a foreground point and 0 indicates a
+            background point.
+          boxes (np.ndarray or None): A Bx4 array given a box prompt to the
+            model, in XYXY format.
+          mask_input (np.ndarray): A low resolution mask input to the model, typically
+            coming from a previous prediction iteration. Has form Bx1xHxW, where
+            for SAM, H=W=256. Masks returned by a previous iteration of the
+            predict method do not need further transformation.
+          multimask_output (bool): If true, the model will return three masks.
+            For ambiguous input prompts (such as a single click), this will often
+            produce better masks than a single prediction. If only a single
+            mask is needed, the model's predicted quality score can be used
+            to select the best mask. For non-ambiguous prompts, such as multiple
+            input prompts, multimask_output=False can give better results.
+          return_logits (bool): If true, returns un-thresholded masks logits
+            instead of a binary mask.
+
+        Returns:
+          (torch.Tensor): The output masks in BxCxHxW format, where C is the
+            number of masks, and (H, W) is the original image size.
+          (torch.Tensor): An array of shape BxC containing the model's
+            predictions for the quality of each mask.
+          (torch.Tensor): An array of shape BxCxHxW, where C is the number
+            of masks and H=W=256. These low res logits can be passed to
+            a subsequent iteration as mask input.
+        """
+        if not self._is_image_set:
+            raise RuntimeError("An image must be set with .set_image(...) before mask prediction.")
+
+        if point_coords is not None:
+            concat_points = (point_coords, point_labels)
+        else:
+            concat_points = None
+
+        # Embed prompts
+        if boxes is not None:
+            box_coords = boxes.reshape(-1, 2, 2)
+            box_labels = torch.tensor([[2, 3]], dtype=torch.int, device=boxes.device)
+            box_labels = box_labels.repeat(boxes.size(0), 1)
+            # we merge "boxes" and "points" into a single "concat_points" input (where
+            # boxes are added at the beginning) to sam_prompt_encoder
+            if concat_points is not None:
+                concat_coords = torch.cat([box_coords, concat_points[0]], dim=1)
+                concat_labels = torch.cat([box_labels, concat_points[1]], dim=1)
+                concat_points = (concat_coords, concat_labels)
+            else:
+                concat_points = (box_coords, box_labels)
+
+        sparse_embeddings, dense_embeddings = self.model.sam_prompt_encoder(
+            points=concat_points,
+            boxes=None,
+            masks=mask_input,
+        )
+
+        # Predict masks
+        batched_mode = (concat_points is not None and concat_points[0].shape[0] > 1)  # multi object prediction
+        high_res_features = [feat_level[img_idx].unsqueeze(0) for feat_level in self._features["high_res_feats"]]
+        low_res_masks, iou_predictions, _, _ = self.model.sam_mask_decoder(
+            image_embeddings=self._features["image_embed"][img_idx].unsqueeze(0),
+            image_pe=self.model.sam_prompt_encoder.get_dense_pe(),
+            sparse_prompt_embeddings=sparse_embeddings,
+            dense_prompt_embeddings=dense_embeddings,
+            multimask_output=multimask_output,
+            repeat_image=batched_mode,
+            high_res_features=high_res_features,
+        )
+
+        # Upscale the masks to the original image resolution
+        masks = self._transforms.postprocess_masks(low_res_masks, self._orig_hw[img_idx])
+        low_res_masks = torch.clamp(low_res_masks, -32.0, 32.0)
+        if not return_logits:
+            masks = masks > self.mask_threshold
+
+        return masks, iou_predictions, low_res_masks
+
+    def get_image_embedding(self) -> torch.Tensor:
+        """
+        Returns the image embeddings for the currently set image, with
+        shape 1xCxHxW, where C is the embedding dimension and (H,W) are
+        the embedding spatial dimension of SAM (typically C=256, H=W=64).
+        """
+        if not self._is_image_set:
+            raise RuntimeError("An image must be set with .set_image(...) to generate an embedding.")
+        assert (self._features is not None), "Features must exist if an image has been set."
+        return self._features["image_embed"]
+
+    @property
+    def device(self) -> torch.device:
+        return self.model.device
+
+    def reset_predictor(self) -> None:
+        """
+        Resets the image embeddings and other state variables.
+        """
+        self._is_image_set = False
+        self._features = None
+        self._orig_hw = None
+        self._is_batch = False

sam2/sam2_train.py

@@ -0,0 +1,575 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.
+
+import logging
+
+import numpy as np
+import torch
+import torch.distributed
+from tensordict import tensorclass
+
+from sam2.modeling.sam2_base import SAM2Base
+from sam2.modeling.sam2_utils import get_next_point, sample_box_points
+from sam2.utils.misc import concat_points
+
+
+@tensorclass
+class BatchedVideoDatapoint:
+    """
+    This class represents a batch of videos with associated annotations.
+    Attributes:
+        img_batch: A [TxBxCxHxW] tensor containing the image data for each frame in the batch, where T is the number of frames per video, and B is the number of videos in the batch.
+        obj_to_frame_idx: A [TxOx2] tensor containing the image_batch index which the object belongs to. O is the number of objects in the batch.
+        masks: A [TxOxHxW] tensor containing binary masks for each object in the batch.
+    """
+
+    img_batch: torch.FloatTensor
+    obj_to_frame_idx: torch.IntTensor
+    masks: torch.BoolTensor
+
+    @property
+    def num_frames(self) -> int:
+        """
+        Returns the number of frames per video.
+        """
+        return self.img_batch.shape[0]
+
+    @property
+    def num_videos(self) -> int:
+        """
+        Returns the number of videos in the batch.
+        """
+        return self.img_batch.shape[1]
+
+    @property
+    def flat_obj_to_img_idx(self) -> torch.IntTensor:
+        """
+        Returns a flattened tensor containing the object to img index.
+        The flat index can be used to access a flattened img_batch of shape [(T*B)xCxHxW]
+        """
+        frame_idx, video_idx = self.obj_to_frame_idx.unbind(dim=-1)
+        flat_idx = video_idx * self.num_frames + frame_idx
+        return flat_idx
+
+    @property
+    def flat_img_batch(self) -> torch.FloatTensor:
+        """
+        Returns a flattened img_batch_tensor of shape [(B*T)xCxHxW]
+        """
+        return self.img_batch.transpose(0, 1).flatten(0, 1)
+
+
+class SAM2Train(SAM2Base):
+
+    def __init__(
+        self,
+        image_encoder,
+        memory_attention=None,
+        memory_encoder=None,
+        prob_to_use_pt_input_for_train=0.0,
+        prob_to_use_pt_input_for_eval=0.0,
+        prob_to_use_box_input_for_train=0.0,
+        prob_to_use_box_input_for_eval=0.0,
+        # if it is greater than 1, we interactive point sampling in the 1st frame and other randomly selected frames
+        num_frames_to_correct_for_train=1,  # default: only iteratively sample on first frame
+        num_frames_to_correct_for_eval=1,  # default: only iteratively sample on first frame
+        rand_frames_to_correct_for_train=False,
+        rand_frames_to_correct_for_eval=False,
+        # how many frames to use as initial conditioning frames (for both point input and mask input; the first frame is always used as an initial conditioning frame)
+        # - if `rand_init_cond_frames` below is True, we randomly sample 1~num_init_cond_frames initial conditioning frames
+        # - otherwise we sample a fixed number of num_init_cond_frames initial conditioning frames
+        # note: for point input, we sample correction points on all such initial conditioning frames, and we require that `num_frames_to_correct` >= `num_init_cond_frames`;
+        # these are initial conditioning frames because as we track the video, more conditioning frames might be added
+        # when a frame receives correction clicks under point input if `add_all_frames_to_correct_as_cond=True`
+        num_init_cond_frames_for_train=1,  # default: only use the first frame as initial conditioning frame
+        num_init_cond_frames_for_eval=1,  # default: only use the first frame as initial conditioning frame
+        rand_init_cond_frames_for_train=True,  # default: random 1~num_init_cond_frames_for_train cond frames (to be constent w/ previous TA data loader)
+        rand_init_cond_frames_for_eval=False,
+        # if `add_all_frames_to_correct_as_cond` is True, we also append to the conditioning frame list any frame that receives a later correction click
+        # if `add_all_frames_to_correct_as_cond` is False, we conditioning frame list to only use those initial conditioning frames
+        add_all_frames_to_correct_as_cond=False,
+        # how many additional correction points to sample (on each frame selected to be corrected)
+        # note that the first frame receives an initial input click (in addition to any correction clicks)
+        num_correction_pt_per_frame=7,
+        # method for point sampling during evaluation
+        # "uniform" (sample uniformly from error region) or "center" (use the point with the largest distance to error region boundary)
+        # default to "center" to be consistent with evaluation in the SAM paper
+        pt_sampling_for_eval="center",
+        # During training, we optionally allow sampling the correction points from GT regions
+        # instead of the prediction error regions with a small probability. This might allow the
+        # model to overfit less to the error regions in training datasets
+        prob_to_sample_from_gt_for_train=0.0,
+        use_act_ckpt_iterative_pt_sampling=False,
+        # whether to forward image features per frame (as it's being tracked) during evaluation, instead of forwarding image features
+        # of all frames at once. This avoids backbone OOM errors on very long videos in evaluation, but could be slightly slower.
+        forward_backbone_per_frame_for_eval=False,
+        freeze_image_encoder=False,
+        **kwargs,
+    ):
+        super().__init__(image_encoder, memory_attention, memory_encoder, **kwargs)
+        self.use_act_ckpt_iterative_pt_sampling = use_act_ckpt_iterative_pt_sampling
+        self.forward_backbone_per_frame_for_eval = forward_backbone_per_frame_for_eval
+
+        # Point sampler and conditioning frames
+        self.prob_to_use_pt_input_for_train = prob_to_use_pt_input_for_train
+        self.prob_to_use_box_input_for_train = prob_to_use_box_input_for_train
+        self.prob_to_use_pt_input_for_eval = prob_to_use_pt_input_for_eval
+        self.prob_to_use_box_input_for_eval = prob_to_use_box_input_for_eval
+        if prob_to_use_pt_input_for_train > 0 or prob_to_use_pt_input_for_eval > 0:
+            logging.info(f"Training with points (sampled from masks) as inputs with p={prob_to_use_pt_input_for_train}")
+            assert num_frames_to_correct_for_train >= num_init_cond_frames_for_train
+            assert num_frames_to_correct_for_eval >= num_init_cond_frames_for_eval
+
+        self.num_frames_to_correct_for_train = num_frames_to_correct_for_train
+        self.num_frames_to_correct_for_eval = num_frames_to_correct_for_eval
+        self.rand_frames_to_correct_for_train = rand_frames_to_correct_for_train
+        self.rand_frames_to_correct_for_eval = rand_frames_to_correct_for_eval
+        # Initial multi-conditioning frames
+        self.num_init_cond_frames_for_train = num_init_cond_frames_for_train
+        self.num_init_cond_frames_for_eval = num_init_cond_frames_for_eval
+        self.rand_init_cond_frames_for_train = rand_init_cond_frames_for_train
+        self.rand_init_cond_frames_for_eval = rand_init_cond_frames_for_eval
+        self.add_all_frames_to_correct_as_cond = add_all_frames_to_correct_as_cond
+        self.num_correction_pt_per_frame = num_correction_pt_per_frame
+        self.pt_sampling_for_eval = pt_sampling_for_eval
+        self.prob_to_sample_from_gt_for_train = prob_to_sample_from_gt_for_train
+        # A random number generator with a fixed initial seed across GPUs
+        self.rng = np.random.default_rng(seed=42)
+
+        if freeze_image_encoder:
+            for p in self.image_encoder.parameters():
+                p.requires_grad = False
+
+    def forward(self, input: BatchedVideoDatapoint, hidden):
+        if self.training or not self.forward_backbone_per_frame_for_eval:
+            # precompute image features on all frames before tracking
+            backbone_out = self.forward_image(input.flat_img_batch)
+        else:
+            # defer image feature computation on a frame until it's being tracked
+            backbone_out = {"backbone_fpn": None, "vision_pos_enc": None}
+        # NOTE: backbone_out = self.prepare_prompt_inputs(backbone_out, input)
+        previous_stages_out = self.forward_tracking(backbone_out, input, hidden)
+
+        return previous_stages_out
+
+    def _prepare_backbone_features_per_frame(self, img_batch, img_ids):
+        """Compute the image backbone features on the fly for the given img_ids."""
+        # Only forward backbone on unique image ids to avoid repetitive computation
+        # (if `img_ids` has only one element, it's already unique so we skip this step).
+        if img_ids.numel() > 1:
+            unique_img_ids, inv_ids = torch.unique(img_ids, return_inverse=True)
+        else:
+            unique_img_ids, inv_ids = img_ids, None
+
+        # Compute the image features on those unique image ids
+        image = img_batch[unique_img_ids]
+        backbone_out = self.forward_image(image)
+        (
+            _,
+            vision_feats,
+            vision_pos_embeds,
+            feat_sizes,
+        ) = self._prepare_backbone_features(backbone_out)
+        # Inverse-map image features for `unique_img_ids` to the final image features
+        # for the original input `img_ids`.
+        if inv_ids is not None:
+            image = image[inv_ids]
+            vision_feats = [x[:, inv_ids] for x in vision_feats]
+            vision_pos_embeds = [x[:, inv_ids] for x in vision_pos_embeds]
+
+        return image, vision_feats, vision_pos_embeds, feat_sizes
+
+    def prepare_prompt_inputs(self, backbone_out, input, start_frame_idx=0):
+        """
+        Prepare input mask, point or box prompts. Optionally, we allow tracking from
+        a custom `start_frame_idx` to the end of the video (for evaluation purposes).
+        """
+        # Load the ground-truth masks on all frames (so that we can later
+        # sample correction points from them)
+        # gt_masks_per_frame = {
+        #     stage_id: targets.segments.unsqueeze(1)  # [B, 1, H_im, W_im]
+        #     for stage_id, targets in enumerate(input.find_targets)
+        # }
+        gt_masks_per_frame = {
+            stage_id: masks.unsqueeze(1)  # [B, 1, H_im, W_im]
+            for stage_id, masks in enumerate(input.masks)
+        }
+        # gt_masks_per_frame = input.masks.unsqueeze(2) # [T,B,1,H_im,W_im] keep everything in tensor form
+        backbone_out["gt_masks_per_frame"] = gt_masks_per_frame
+        num_frames = input.num_frames
+        backbone_out["num_frames"] = num_frames
+
+        # Randomly decide whether to use point inputs or mask inputs
+        if self.training:
+            prob_to_use_pt_input = self.prob_to_use_pt_input_for_train
+            prob_to_use_box_input = self.prob_to_use_box_input_for_train
+            num_frames_to_correct = self.num_frames_to_correct_for_train
+            rand_frames_to_correct = self.rand_frames_to_correct_for_train
+            num_init_cond_frames = self.num_init_cond_frames_for_train
+            rand_init_cond_frames = self.rand_init_cond_frames_for_train
+        else:
+            prob_to_use_pt_input = self.prob_to_use_pt_input_for_eval
+            prob_to_use_box_input = self.prob_to_use_box_input_for_eval
+            num_frames_to_correct = self.num_frames_to_correct_for_eval
+            rand_frames_to_correct = self.rand_frames_to_correct_for_eval
+            num_init_cond_frames = self.num_init_cond_frames_for_eval
+            rand_init_cond_frames = self.rand_init_cond_frames_for_eval
+        if num_frames == 1:
+            # here we handle a special case for mixing video + SAM on image training,
+            # where we force using point input for the SAM task on static images
+            prob_to_use_pt_input = 1.0
+            num_frames_to_correct = 1
+            num_init_cond_frames = 1
+        assert num_init_cond_frames >= 1
+        # (here `self.rng.random()` returns value in range 0.0 <= X < 1.0)
+        use_pt_input = self.rng.random() < prob_to_use_pt_input
+        if rand_init_cond_frames and num_init_cond_frames > 1:
+            # randomly select 1 to `num_init_cond_frames` frames as initial conditioning frames
+            num_init_cond_frames = self.rng.integers(1, num_init_cond_frames, endpoint=True)
+        if (use_pt_input and rand_frames_to_correct and num_frames_to_correct > num_init_cond_frames):
+            # randomly select `num_init_cond_frames` to `num_frames_to_correct` frames to sample
+            # correction clicks (only for the case of point input)
+            num_frames_to_correct = self.rng.integers(num_init_cond_frames, num_frames_to_correct, endpoint=True)
+        backbone_out["use_pt_input"] = use_pt_input
+
+        # Sample initial conditioning frames
+        if num_init_cond_frames == 1:
+            init_cond_frames = [start_frame_idx]  # starting frame
+        else:
+            # starting frame + randomly selected remaining frames (without replacement)
+            init_cond_frames = [start_frame_idx] + self.rng.choice(
+                range(start_frame_idx + 1, num_frames),
+                num_init_cond_frames - 1,
+                replace=False,
+            ).tolist()
+        backbone_out["init_cond_frames"] = init_cond_frames
+        backbone_out["frames_not_in_init_cond"] = [
+            t for t in range(start_frame_idx, num_frames) if t not in init_cond_frames
+        ]
+        # Prepare mask or point inputs on initial conditioning frames
+        backbone_out["mask_inputs_per_frame"] = {}  # {frame_idx: <input_masks>}
+        backbone_out["point_inputs_per_frame"] = {}  # {frame_idx: <input_points>}
+        for t in init_cond_frames:
+            if not use_pt_input:
+                backbone_out["mask_inputs_per_frame"][t] = gt_masks_per_frame[t]
+            else:
+                # During training # P(box) = prob_to_use_pt_input * prob_to_use_box_input
+                use_box_input = self.rng.random() < prob_to_use_box_input
+                if use_box_input:
+                    points, labels = sample_box_points(gt_masks_per_frame[t], )
+                else:
+                    # (here we only sample **one initial point** on initial conditioning frames from the
+                    # ground-truth mask; we may sample more correction points on the fly)
+                    points, labels = get_next_point(
+                        gt_masks=gt_masks_per_frame[t],
+                        pred_masks=None,
+                        method=("uniform" if self.training else self.pt_sampling_for_eval),
+                    )
+
+                point_inputs = {"point_coords": points, "point_labels": labels}
+                backbone_out["point_inputs_per_frame"][t] = point_inputs
+
+        # Sample frames where we will add correction clicks on the fly
+        # based on the error between prediction and ground-truth masks
+        if not use_pt_input:
+            # no correction points will be sampled when using mask inputs
+            frames_to_add_correction_pt = []
+        elif num_frames_to_correct == num_init_cond_frames:
+            frames_to_add_correction_pt = init_cond_frames
+        else:
+            assert num_frames_to_correct > num_init_cond_frames
+            # initial cond frame + randomly selected remaining frames (without replacement)
+            extra_num = num_frames_to_correct - num_init_cond_frames
+            frames_to_add_correction_pt = (
+                init_cond_frames +
+                self.rng.choice(backbone_out["frames_not_in_init_cond"], extra_num, replace=False).tolist())
+        backbone_out["frames_to_add_correction_pt"] = frames_to_add_correction_pt
+
+        return backbone_out
+
+    def forward_tracking(self, backbone_out, input: BatchedVideoDatapoint, hidden, return_dict=False):
+        """Forward video tracking on each frame (and sample correction clicks)."""
+        img_feats_already_computed = backbone_out["backbone_fpn"] is not None
+        if img_feats_already_computed:
+            # Prepare the backbone features
+            # - vision_feats and vision_pos_embeds are in (HW)BC format
+            (
+                _,
+                vision_feats,
+                vision_pos_embeds,
+                feat_sizes,
+            ) = self._prepare_backbone_features(backbone_out)
+
+        # Starting the stage loop
+        # NOTE: num_frames = backbone_out["num_frames"] =========================================
+        num_frames = input.num_frames
+        # =======================================================================================
+        # NOTE: init_cond_frames = backbone_out["init_cond_frames"] =============================
+        # init_cond_frames = list(range(num_frames))
+        init_cond_frames = [0]
+        # =======================================================================================
+        # NOTE: frames_to_add_correction_pt = backbone_out["frames_to_add_correction_pt"] =======
+        frames_to_add_correction_pt = []
+        # =======================================================================================
+        # first process all the initial conditioning frames to encode them as memory,
+        # and then conditioning on them to track the remaining frames
+        # NOTE: processing_order = init_cond_frames + backbone_out["frames_not_in_init_cond"] ===
+        frames_not_in_init_cond = [t for t in range(num_frames) if t not in init_cond_frames]
+        processing_order = init_cond_frames + frames_not_in_init_cond
+        # =======================================================================================
+        backbone_out["point_inputs_per_frame"] = {}
+        backbone_out["mask_inputs_per_frame"] = {}
+        # backbone_out["hidden_inputs_per_frame"] = {stage_id: hidden for stage_id in processing_order}
+        backbone_out["hidden_inputs_per_frame"] = {0: hidden}
+        backbone_out["gt_masks_per_frame"] = {
+            stage_id: masks.unsqueeze(1)  # [B, 1, H_im, W_im]
+            for stage_id, masks in enumerate(input.masks)
+        }
+        # =======================================================================================
+        output_dict = {
+            "cond_frame_outputs": {},  # dict containing {frame_idx: <out>}
+            "non_cond_frame_outputs": {},  # dict containing {frame_idx: <out>}
+        }
+        for stage_id in processing_order:
+            # Get the image features for the current frames
+            # img_ids = input.find_inputs[stage_id].img_ids
+            img_ids = input.flat_obj_to_img_idx[stage_id]
+            if img_feats_already_computed:
+                # Retrieve image features according to img_ids (if they are already computed).
+                current_vision_feats = [x[:, img_ids] for x in vision_feats]
+                current_vision_pos_embeds = [x[:, img_ids] for x in vision_pos_embeds]
+            else:
+                # Otherwise, compute the image features on the fly for the given img_ids
+                # (this might be used for evaluation on long videos to avoid backbone OOM).
+                (
+                    _,
+                    current_vision_feats,
+                    current_vision_pos_embeds,
+                    feat_sizes,
+                ) = self._prepare_backbone_features_per_frame(input.flat_img_batch, img_ids)
+
+            # Get output masks based on this frame's prompts and previous memory
+            current_out = self.track_step(
+                frame_idx=stage_id,
+                is_init_cond_frame=stage_id in init_cond_frames,
+                current_vision_feats=current_vision_feats,
+                current_vision_pos_embeds=current_vision_pos_embeds,
+                feat_sizes=feat_sizes,
+                point_inputs=backbone_out["point_inputs_per_frame"].get(stage_id, None),
+                mask_inputs=backbone_out["mask_inputs_per_frame"].get(stage_id, None),
+                hidden_inputs=backbone_out["hidden_inputs_per_frame"].get(stage_id, None),
+                gt_masks=backbone_out["gt_masks_per_frame"].get(stage_id, None),
+                frames_to_add_correction_pt=frames_to_add_correction_pt,
+                output_dict=output_dict,
+                num_frames=num_frames,
+            )
+            # Append the output, depending on whether it's a conditioning frame
+            add_output_as_cond_frame = stage_id in init_cond_frames or (self.add_all_frames_to_correct_as_cond
+                                                                        and stage_id in frames_to_add_correction_pt)
+            if add_output_as_cond_frame:
+                output_dict["cond_frame_outputs"][stage_id] = current_out
+            else:
+                output_dict["non_cond_frame_outputs"][stage_id] = current_out
+
+        if return_dict:
+            return output_dict
+        # turn `output_dict` into a list for loss function
+        all_frame_outputs = {}
+        all_frame_outputs.update(output_dict["cond_frame_outputs"])
+        all_frame_outputs.update(output_dict["non_cond_frame_outputs"])
+        all_frame_outputs = [all_frame_outputs[t] for t in range(num_frames)]
+        # Make DDP happy with activation checkpointing by removing unused keys
+        all_frame_outputs = [{k: v for k, v in d.items() if k != "obj_ptr"} for d in all_frame_outputs]
+
+        return all_frame_outputs
+
+    def track_step(
+        self,
+        frame_idx,
+        is_init_cond_frame,
+        current_vision_feats,
+        current_vision_pos_embeds,
+        feat_sizes,
+        point_inputs,
+        mask_inputs,
+        hidden_inputs,
+        output_dict,
+        num_frames,
+        track_in_reverse=False,  # tracking in reverse time order (for demo usage)
+        run_mem_encoder=True,  # Whether to run the memory encoder on the predicted masks.
+        prev_sam_mask_logits=None,  # The previously predicted SAM mask logits.
+        frames_to_add_correction_pt=None,
+        gt_masks=None,
+    ):
+        if frames_to_add_correction_pt is None:
+            frames_to_add_correction_pt = []
+        current_out, sam_outputs, high_res_features, pix_feat = self._track_step(
+            frame_idx,
+            is_init_cond_frame,
+            current_vision_feats,
+            current_vision_pos_embeds,
+            feat_sizes,
+            point_inputs,
+            mask_inputs,
+            hidden_inputs,
+            output_dict,
+            num_frames,
+            track_in_reverse,
+            prev_sam_mask_logits,
+        )
+
+        (
+            low_res_multimasks,
+            high_res_multimasks,
+            ious,
+            low_res_masks,
+            high_res_masks,
+            obj_ptr,
+            object_score_logits,
+        ) = sam_outputs
+
+        current_out["multistep_pred_masks"] = low_res_masks
+        current_out["multistep_pred_masks_high_res"] = high_res_masks
+        current_out["multistep_pred_multimasks"] = [low_res_multimasks]
+        current_out["multistep_pred_multimasks_high_res"] = [high_res_multimasks]
+        current_out["multistep_pred_ious"] = [ious]
+        current_out["multistep_point_inputs"] = [point_inputs]
+        current_out["multistep_object_score_logits"] = [object_score_logits]
+
+        # Optionally, sample correction points iteratively to correct the mask
+        if frame_idx in frames_to_add_correction_pt:
+            point_inputs, final_sam_outputs = self._iter_correct_pt_sampling(
+                is_init_cond_frame,
+                point_inputs,
+                gt_masks,
+                high_res_features,
+                pix_feat,
+                low_res_multimasks,
+                high_res_multimasks,
+                ious,
+                low_res_masks,
+                high_res_masks,
+                object_score_logits,
+                current_out,
+            )
+            (
+                _,
+                _,
+                _,
+                low_res_masks,
+                high_res_masks,
+                obj_ptr,
+                object_score_logits,
+            ) = final_sam_outputs
+
+        # Use the final prediction (after all correction steps for output and eval)
+        current_out["pred_masks"] = low_res_masks
+        current_out["pred_masks_high_res"] = high_res_masks
+        current_out["obj_ptr"] = obj_ptr
+
+        # Finally run the memory encoder on the predicted mask to encode
+        # it into a new memory feature (that can be used in future frames)
+        self._encode_memory_in_output(
+            current_vision_feats,
+            feat_sizes,
+            point_inputs,
+            run_mem_encoder,
+            high_res_masks,
+            object_score_logits,
+            current_out,
+        )
+        return current_out
+
+    def _iter_correct_pt_sampling(
+        self,
+        is_init_cond_frame,
+        point_inputs,
+        gt_masks,
+        high_res_features,
+        pix_feat_with_mem,
+        low_res_multimasks,
+        high_res_multimasks,
+        ious,
+        low_res_masks,
+        high_res_masks,
+        object_score_logits,
+        current_out,
+    ):
+
+        assert gt_masks is not None
+        all_pred_masks = [low_res_masks]
+        all_pred_high_res_masks = [high_res_masks]
+        all_pred_multimasks = [low_res_multimasks]
+        all_pred_high_res_multimasks = [high_res_multimasks]
+        all_pred_ious = [ious]
+        all_point_inputs = [point_inputs]
+        all_object_score_logits = [object_score_logits]
+        for _ in range(self.num_correction_pt_per_frame):
+            # sample a new point from the error between prediction and ground-truth
+            # (with a small probability, directly sample from GT masks instead of errors)
+            if self.training and self.prob_to_sample_from_gt_for_train > 0:
+                sample_from_gt = (self.rng.random() < self.prob_to_sample_from_gt_for_train)
+            else:
+                sample_from_gt = False
+            # if `pred_for_new_pt` is None, only GT masks will be used for point sampling
+            pred_for_new_pt = None if sample_from_gt else (high_res_masks > 0)
+            new_points, new_labels = get_next_point(
+                gt_masks=gt_masks,
+                pred_masks=pred_for_new_pt,
+                method="uniform" if self.training else self.pt_sampling_for_eval,
+            )
+            point_inputs = concat_points(point_inputs, new_points, new_labels)
+            # Feed the mask logits of the previous SAM outputs in the next SAM decoder step.
+            # For tracking, this means that when the user adds a correction click, we also feed
+            # the tracking output mask logits along with the click as input to the SAM decoder.
+            mask_inputs = low_res_masks
+            multimask_output = self._use_multimask(is_init_cond_frame, point_inputs)
+            if self.use_act_ckpt_iterative_pt_sampling and not multimask_output:
+                sam_outputs = torch.utils.checkpoint.checkpoint(
+                    self._forward_sam_heads,
+                    backbone_features=pix_feat_with_mem,
+                    point_inputs=point_inputs,
+                    mask_inputs=mask_inputs,
+                    high_res_features=high_res_features,
+                    multimask_output=multimask_output,
+                    use_reentrant=False,
+                )
+            else:
+                sam_outputs = self._forward_sam_heads(
+                    backbone_features=pix_feat_with_mem,
+                    point_inputs=point_inputs,
+                    mask_inputs=mask_inputs,
+                    high_res_features=high_res_features,
+                    multimask_output=multimask_output,
+                )
+            (
+                low_res_multimasks,
+                high_res_multimasks,
+                ious,
+                low_res_masks,
+                high_res_masks,
+                _,
+                object_score_logits,
+            ) = sam_outputs
+            all_pred_masks.append(low_res_masks)
+            all_pred_high_res_masks.append(high_res_masks)
+            all_pred_multimasks.append(low_res_multimasks)
+            all_pred_high_res_multimasks.append(high_res_multimasks)
+            all_pred_ious.append(ious)
+            all_point_inputs.append(point_inputs)
+            all_object_score_logits.append(object_score_logits)
+
+        # Concatenate the masks along channel (to compute losses on all of them,
+        # using `MultiStepIteractiveMasks`)
+        current_out["multistep_pred_masks"] = torch.cat(all_pred_masks, dim=1)
+        current_out["multistep_pred_masks_high_res"] = torch.cat(all_pred_high_res_masks, dim=1)
+        current_out["multistep_pred_multimasks"] = all_pred_multimasks
+        current_out["multistep_pred_multimasks_high_res"] = all_pred_high_res_multimasks
+        current_out["multistep_pred_ious"] = all_pred_ious
+        current_out["multistep_point_inputs"] = all_point_inputs
+        current_out["multistep_object_score_logits"] = all_object_score_logits
+
+        return point_inputs, sam_outputs

sam2/sam2_video_predictor.py

@@ -0,0 +1,1272 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.
+
+from collections import OrderedDict
+
+import torch
+import torch.nn.functional as F
+from tqdm import tqdm
+
+from sam2.modeling.sam2_base import NO_OBJ_SCORE, SAM2Base
+from sam2.utils.misc import concat_points, fill_holes_in_mask_scores, load_video_frames
+
+
+class SAM2VideoPredictor(SAM2Base):
+    """The predictor class to handle user interactions and manage inference states."""
+
+    def __init__(
+        self,
+        fill_hole_area=0,
+        # whether to apply non-overlapping constraints on the output object masks
+        non_overlap_masks=False,
+        # whether to clear non-conditioning memory of the surrounding frames (which may contain outdated information) after adding correction clicks;
+        # note that this would only apply to *single-object tracking* unless `clear_non_cond_mem_for_multi_obj` is also set to True)
+        clear_non_cond_mem_around_input=False,
+        # if `add_all_frames_to_correct_as_cond` is True, we also append to the conditioning frame list any frame that receives a later correction click
+        # if `add_all_frames_to_correct_as_cond` is False, we conditioning frame list to only use those initial conditioning frames
+        add_all_frames_to_correct_as_cond=False,
+        inference_mode=True,
+        **kwargs,
+    ):
+        super().__init__(**kwargs)
+        self.fill_hole_area = fill_hole_area
+        self.non_overlap_masks = non_overlap_masks
+        self.clear_non_cond_mem_around_input = clear_non_cond_mem_around_input
+        self.add_all_frames_to_correct_as_cond = add_all_frames_to_correct_as_cond
+        self.inference_mode = inference_mode
+
+    @property
+    def dtype(self):
+        return self.image_encoder.trunk.patch_embed.proj.weight.dtype
+
+    def init_state(
+        self,
+        frame,
+        frame_size=None,
+        offload_video_to_cpu=False,
+        offload_state_to_cpu=False,
+        async_loading_frames=False,
+    ):
+        """Initialize an inference state."""
+        compute_device = self.device  # device of the model
+        if isinstance(frame, str):
+            images, video_height, video_width = load_video_frames(
+                video_path=frame,
+                image_size=self.image_size,
+                offload_video_to_cpu=offload_video_to_cpu,
+                async_loading_frames=async_loading_frames,
+                compute_device=compute_device,
+            )
+        else:
+            if frame_size is None:
+                frame_size = (self.image_size, self.image_size)
+            images, video_height, video_width = (frame, *frame_size)
+        inference_state = {}
+        inference_state["images"] = images
+        inference_state["num_frames"] = len(images)
+        # whether to offload the video frames to CPU memory
+        # turning on this option saves the GPU memory with only a very small overhead
+        inference_state["offload_video_to_cpu"] = offload_video_to_cpu
+        # whether to offload the inference state to CPU memory
+        # turning on this option saves the GPU memory at the cost of a lower tracking fps
+        # (e.g. in a test case of 768x768 model, fps dropped from 27 to 24 when tracking one object
+        # and from 24 to 21 when tracking two objects)
+        inference_state["offload_state_to_cpu"] = offload_state_to_cpu
+        # the original video height and width, used for resizing final output scores
+        inference_state["video_height"] = video_height
+        inference_state["video_width"] = video_width
+        inference_state["device"] = compute_device
+        if offload_state_to_cpu:
+            inference_state["storage_device"] = torch.device("cpu")
+        else:
+            inference_state["storage_device"] = compute_device
+        # inputs on each frame
+        inference_state["point_inputs_per_obj"] = {}
+        inference_state["mask_inputs_per_obj"] = {}
+        # visual features on a small number of recently visited frames for quick interactions
+        inference_state["cached_features"] = {}
+        # values that don't change across frames (so we only need to hold one copy of them)
+        inference_state["constants"] = {}
+        # mapping between client-side object id and model-side object index
+        inference_state["obj_id_to_idx"] = OrderedDict()
+        inference_state["obj_idx_to_id"] = OrderedDict()
+        inference_state["obj_ids"] = []
+        # Slice (view) of each object tracking results, sharing the same memory with "output_dict"
+        inference_state["output_dict_per_obj"] = {}
+        # A temporary storage to hold new outputs when user interact with a frame
+        # to add clicks or mask (it's merged into "output_dict" before propagation starts)
+        inference_state["temp_output_dict_per_obj"] = {}
+        # Frames that already holds consolidated outputs from click or mask inputs
+        # (we directly use their consolidated outputs during tracking)
+        # metadata for each tracking frame (e.g. which direction it's tracked)
+        inference_state["frames_tracked_per_obj"] = {}
+        # Warm up the visual backbone and cache the image feature on all frames
+        self._get_image_feature(inference_state, frame_idx=0, batch_size=1)
+        return inference_state
+
+    @classmethod
+    def from_pretrained(cls, model_id: str, **kwargs) -> "SAM2VideoPredictor":
+        """
+        Load a pretrained model from the Hugging Face hub.
+
+        Arguments:
+          model_id (str): The Hugging Face repository ID.
+          **kwargs: Additional arguments to pass to the model constructor.
+
+        Returns:
+          (SAM2VideoPredictor): The loaded model.
+        """
+        from sam2.build_sam import build_sam2_video_predictor_hf
+
+        sam_model = build_sam2_video_predictor_hf(model_id, **kwargs)
+        return sam_model
+
+    def _obj_id_to_idx(self, inference_state, obj_id):
+        """Map client-side object id to model-side object index."""
+        obj_idx = inference_state["obj_id_to_idx"].get(obj_id, None)
+        if obj_idx is not None:
+            return obj_idx
+
+        # We always allow adding new objects (including after tracking starts)
+        # get the next object slot
+        obj_idx = len(inference_state["obj_id_to_idx"])
+        inference_state["obj_id_to_idx"][obj_id] = obj_idx
+        inference_state["obj_idx_to_id"][obj_idx] = obj_id
+        inference_state["obj_ids"] = list(inference_state["obj_id_to_idx"])
+        # set up input and output structures for this object
+        inference_state["point_inputs_per_obj"][obj_idx] = {}
+        inference_state["mask_inputs_per_obj"][obj_idx] = {}
+        inference_state["output_dict_per_obj"][obj_idx] = {
+            "cond_frame_outputs": {},  # dict containing {frame_idx: <out>}
+            "non_cond_frame_outputs": {},  # dict containing {frame_idx: <out>}
+        }
+        inference_state["temp_output_dict_per_obj"][obj_idx] = {
+            "cond_frame_outputs": {},  # dict containing {frame_idx: <out>}
+            "non_cond_frame_outputs": {},  # dict containing {frame_idx: <out>}
+        }
+        inference_state["frames_tracked_per_obj"][obj_idx] = {}
+        return obj_idx
+
+    def _obj_idx_to_id(self, inference_state, obj_idx):
+        """Map model-side object index to client-side object id."""
+        return inference_state["obj_idx_to_id"][obj_idx]
+
+    def _get_obj_num(self, inference_state):
+        """Get the total number of unique object ids received so far in this session."""
+        return len(inference_state["obj_idx_to_id"])
+
+    @torch.inference_mode()
+    def add_new_hidden_state(
+        self,
+        inference_state,
+        frame_idx,
+        obj_id,
+        hidden,
+    ):
+        obj_idx = self._obj_id_to_idx(inference_state, obj_id)
+        # If this frame hasn't been tracked before, we treat it as an initial conditioning
+        # frame, meaning that the inputs points are to generate segments on this frame without
+        # using any memory from other frames, like in SAM. Otherwise (if it has been tracked),
+        # the input points will be used to correct the already tracked masks.
+        obj_frames_tracked = inference_state["frames_tracked_per_obj"][obj_idx]
+        is_init_cond_frame = frame_idx not in obj_frames_tracked
+        # whether to track in reverse time order
+        if is_init_cond_frame:
+            reverse = False
+        else:
+            reverse = obj_frames_tracked[frame_idx]["reverse"]
+        obj_output_dict = inference_state["output_dict_per_obj"][obj_idx]
+        obj_temp_output_dict = inference_state["temp_output_dict_per_obj"][obj_idx]
+        # Add a frame to conditioning output if it's an initial conditioning frame or
+        # if the model sees all frames receiving clicks/mask as conditioning frames.
+        is_cond = is_init_cond_frame or self.add_all_frames_to_correct_as_cond
+        storage_key = "cond_frame_outputs" if is_cond else "non_cond_frame_outputs"
+
+        # Get any previously predicted mask logits on this object and feed it along with
+        # the new clicks into the SAM mask decoder.
+        prev_sam_mask_logits = None
+        # lookup temporary output dict first, which contains the most recent output
+        # (if not found, then lookup conditioning and non-conditioning frame output)
+        prev_out = obj_temp_output_dict[storage_key].get(frame_idx)
+        if prev_out is None:
+            prev_out = obj_output_dict["cond_frame_outputs"].get(frame_idx)
+            if prev_out is None:
+                prev_out = obj_output_dict["non_cond_frame_outputs"].get(frame_idx)
+
+        if prev_out is not None and prev_out["pred_masks"] is not None:
+            device = inference_state["device"]
+            prev_sam_mask_logits = prev_out["pred_masks"].to(device, non_blocking=True)
+            # Clamp the scale of prev_sam_mask_logits to avoid rare numerical issues.
+            prev_sam_mask_logits = torch.clamp(prev_sam_mask_logits, -32.0, 32.0)
+        current_out, _ = self._run_single_frame_inference(
+            inference_state=inference_state,
+            output_dict=obj_output_dict,  # run on the slice of a single object
+            frame_idx=frame_idx,
+            batch_size=1,  # run on the slice of a single object
+            is_init_cond_frame=is_init_cond_frame,
+            point_inputs=None,
+            mask_inputs=None,
+            hidden_inputs=hidden,
+            reverse=reverse,
+            # Skip the memory encoder when adding clicks or mask. We execute the memory encoder
+            # at the beginning of `propagate_in_video` (after user finalize their clicks). This
+            # allows us to enforce non-overlapping constraints on all objects before encoding
+            # them into memory.
+            run_mem_encoder=False,
+            prev_sam_mask_logits=prev_sam_mask_logits,
+        )
+        # Add the output to the output dict (to be used as future memory)
+        obj_temp_output_dict[storage_key][frame_idx] = current_out
+
+        # Resize the output mask to the original video resolution
+        obj_ids = inference_state["obj_ids"]
+        consolidated_out = self._consolidate_temp_output_across_obj(
+            inference_state,
+            frame_idx,
+            is_cond=is_cond,
+            consolidate_at_video_res=True,
+        )
+        _, video_res_masks = self._get_orig_video_res_output(inference_state, consolidated_out["pred_masks_video_res"])
+        return frame_idx, obj_ids, video_res_masks
+
+    @torch.inference_mode()
+    def add_new_points_or_box(
+        self,
+        inference_state,
+        frame_idx,
+        obj_id,
+        points=None,
+        labels=None,
+        clear_old_points=True,
+        normalize_coords=True,
+        box=None,
+    ):
+        """Add new points to a frame."""
+        obj_idx = self._obj_id_to_idx(inference_state, obj_id)
+        point_inputs_per_frame = inference_state["point_inputs_per_obj"][obj_idx]
+        mask_inputs_per_frame = inference_state["mask_inputs_per_obj"][obj_idx]
+
+        if (points is not None) != (labels is not None):
+            raise ValueError("points and labels must be provided together")
+        if points is None and box is None:
+            raise ValueError("at least one of points or box must be provided as input")
+
+        if points is None:
+            points = torch.zeros(0, 2, dtype=torch.float32)
+        elif not isinstance(points, torch.Tensor):
+            points = torch.tensor(points, dtype=torch.float32)
+        if labels is None:
+            labels = torch.zeros(0, dtype=torch.int32)
+        elif not isinstance(labels, torch.Tensor):
+            labels = torch.tensor(labels, dtype=torch.int32)
+        if points.dim() == 2:
+            points = points.unsqueeze(0)  # add batch dimension
+        if labels.dim() == 1:
+            labels = labels.unsqueeze(0)  # add batch dimension
+
+        # If `box` is provided, we add it as the first two points with labels 2 and 3
+        # along with the user-provided points (consistent with how SAM 2 is trained).
+        if box is not None:
+            if not clear_old_points:
+                raise ValueError("cannot add box without clearing old points, since "
+                                 "box prompt must be provided before any point prompt "
+                                 "(please use clear_old_points=True instead)")
+            if not isinstance(box, torch.Tensor):
+                box = torch.tensor(box, dtype=torch.float32, device=points.device)
+            box_coords = box.reshape(1, 2, 2)
+            box_labels = torch.tensor([2, 3], dtype=torch.int32, device=labels.device)
+            box_labels = box_labels.reshape(1, 2)
+            points = torch.cat([box_coords, points], dim=1)
+            labels = torch.cat([box_labels, labels], dim=1)
+
+        if normalize_coords:
+            video_H = inference_state["video_height"]
+            video_W = inference_state["video_width"]
+            points = points / torch.tensor([video_W, video_H]).to(points.device)
+        # scale the (normalized) coordinates by the model's internal image size
+        points = points * self.image_size
+        points = points.to(inference_state["device"])
+        labels = labels.to(inference_state["device"])
+
+        if not clear_old_points:
+            point_inputs = point_inputs_per_frame.get(frame_idx, None)
+        else:
+            point_inputs = None
+        point_inputs = concat_points(point_inputs, points, labels)
+
+        point_inputs_per_frame[frame_idx] = point_inputs
+        mask_inputs_per_frame.pop(frame_idx, None)
+        # If this frame hasn't been tracked before, we treat it as an initial conditioning
+        # frame, meaning that the inputs points are to generate segments on this frame without
+        # using any memory from other frames, like in SAM. Otherwise (if it has been tracked),
+        # the input points will be used to correct the already tracked masks.
+        obj_frames_tracked = inference_state["frames_tracked_per_obj"][obj_idx]
+        is_init_cond_frame = frame_idx not in obj_frames_tracked
+        # whether to track in reverse time order
+        if is_init_cond_frame:
+            reverse = False
+        else:
+            reverse = obj_frames_tracked[frame_idx]["reverse"]
+        obj_output_dict = inference_state["output_dict_per_obj"][obj_idx]
+        obj_temp_output_dict = inference_state["temp_output_dict_per_obj"][obj_idx]
+        # Add a frame to conditioning output if it's an initial conditioning frame or
+        # if the model sees all frames receiving clicks/mask as conditioning frames.
+        is_cond = is_init_cond_frame or self.add_all_frames_to_correct_as_cond
+        storage_key = "cond_frame_outputs" if is_cond else "non_cond_frame_outputs"
+
+        # Get any previously predicted mask logits on this object and feed it along with
+        # the new clicks into the SAM mask decoder.
+        prev_sam_mask_logits = None
+        # lookup temporary output dict first, which contains the most recent output
+        # (if not found, then lookup conditioning and non-conditioning frame output)
+        prev_out = obj_temp_output_dict[storage_key].get(frame_idx)
+        if prev_out is None:
+            prev_out = obj_output_dict["cond_frame_outputs"].get(frame_idx)
+            if prev_out is None:
+                prev_out = obj_output_dict["non_cond_frame_outputs"].get(frame_idx)
+
+        if prev_out is not None and prev_out["pred_masks"] is not None:
+            device = inference_state["device"]
+            prev_sam_mask_logits = prev_out["pred_masks"].to(device, non_blocking=True)
+            # Clamp the scale of prev_sam_mask_logits to avoid rare numerical issues.
+            prev_sam_mask_logits = torch.clamp(prev_sam_mask_logits, -32.0, 32.0)
+        current_out, _ = self._run_single_frame_inference(
+            inference_state=inference_state,
+            output_dict=obj_output_dict,  # run on the slice of a single object
+            frame_idx=frame_idx,
+            batch_size=1,  # run on the slice of a single object
+            is_init_cond_frame=is_init_cond_frame,
+            point_inputs=point_inputs,
+            mask_inputs=None,
+            hidden_inputs=None,
+            reverse=reverse,
+            # Skip the memory encoder when adding clicks or mask. We execute the memory encoder
+            # at the beginning of `propagate_in_video` (after user finalize their clicks). This
+            # allows us to enforce non-overlapping constraints on all objects before encoding
+            # them into memory.
+            run_mem_encoder=False,
+            prev_sam_mask_logits=prev_sam_mask_logits,
+        )
+        # Add the output to the output dict (to be used as future memory)
+        obj_temp_output_dict[storage_key][frame_idx] = current_out
+
+        # Resize the output mask to the original video resolution
+        obj_ids = inference_state["obj_ids"]
+        consolidated_out = self._consolidate_temp_output_across_obj(
+            inference_state,
+            frame_idx,
+            is_cond=is_cond,
+            consolidate_at_video_res=True,
+        )
+        _, video_res_masks = self._get_orig_video_res_output(inference_state, consolidated_out["pred_masks_video_res"])
+        return frame_idx, obj_ids, video_res_masks
+
+    def add_new_points(self, *args, **kwargs):
+        """Deprecated method. Please use `add_new_points_or_box` instead."""
+        return self.add_new_points_or_box(*args, **kwargs)
+
+    @torch.inference_mode()
+    def add_new_mask(
+        self,
+        inference_state,
+        frame_idx,
+        obj_id,
+        mask,
+    ):
+        """Add new mask to a frame."""
+        obj_idx = self._obj_id_to_idx(inference_state, obj_id)
+        point_inputs_per_frame = inference_state["point_inputs_per_obj"][obj_idx]
+        mask_inputs_per_frame = inference_state["mask_inputs_per_obj"][obj_idx]
+
+        if not isinstance(mask, torch.Tensor):
+            mask = torch.tensor(mask, dtype=torch.bool)
+        assert mask.dim() == 2
+        mask_H, mask_W = mask.shape
+        mask_inputs_orig = mask[None, None]  # add batch and channel dimension
+        mask_inputs_orig = mask_inputs_orig.float().to(inference_state["device"])
+
+        # resize the mask if it doesn't match the model's image size
+        if mask_H != self.image_size or mask_W != self.image_size:
+            mask_inputs = torch.nn.functional.interpolate(
+                mask_inputs_orig,
+                size=(self.image_size, self.image_size),
+                align_corners=False,
+                mode="bilinear",
+                antialias=True,  # use antialias for downsampling
+            )
+            mask_inputs = (mask_inputs >= 0.5).float()
+        else:
+            mask_inputs = mask_inputs_orig
+
+        mask_inputs_per_frame[frame_idx] = mask_inputs
+        point_inputs_per_frame.pop(frame_idx, None)
+        # If this frame hasn't been tracked before, we treat it as an initial conditioning
+        # frame, meaning that the inputs points are to generate segments on this frame without
+        # using any memory from other frames, like in SAM. Otherwise (if it has been tracked),
+        # the input points will be used to correct the already tracked masks.
+        obj_frames_tracked = inference_state["frames_tracked_per_obj"][obj_idx]
+        is_init_cond_frame = frame_idx not in obj_frames_tracked
+        # whether to track in reverse time order
+        if is_init_cond_frame:
+            reverse = False
+        else:
+            reverse = obj_frames_tracked[frame_idx]["reverse"]
+        obj_output_dict = inference_state["output_dict_per_obj"][obj_idx]
+        obj_temp_output_dict = inference_state["temp_output_dict_per_obj"][obj_idx]
+        # Add a frame to conditioning output if it's an initial conditioning frame or
+        # if the model sees all frames receiving clicks/mask as conditioning frames.
+        is_cond = is_init_cond_frame or self.add_all_frames_to_correct_as_cond
+        storage_key = "cond_frame_outputs" if is_cond else "non_cond_frame_outputs"
+
+        current_out, _ = self._run_single_frame_inference(
+            inference_state=inference_state,
+            output_dict=obj_output_dict,  # run on the slice of a single object
+            frame_idx=frame_idx,
+            batch_size=1,  # run on the slice of a single object
+            is_init_cond_frame=is_init_cond_frame,
+            point_inputs=None,
+            mask_inputs=mask_inputs,
+            hidden_inputs=None,
+            reverse=reverse,
+            # Skip the memory encoder when adding clicks or mask. We execute the memory encoder
+            # at the beginning of `propagate_in_video` (after user finalize their clicks). This
+            # allows us to enforce non-overlapping constraints on all objects before encoding
+            # them into memory.
+            run_mem_encoder=False,
+        )
+        # Add the output to the output dict (to be used as future memory)
+        obj_temp_output_dict[storage_key][frame_idx] = current_out
+
+        # Resize the output mask to the original video resolution
+        obj_ids = inference_state["obj_ids"]
+        consolidated_out = self._consolidate_temp_output_across_obj(
+            inference_state,
+            frame_idx,
+            is_cond=is_cond,
+            consolidate_at_video_res=True,
+        )
+        _, video_res_masks = self._get_orig_video_res_output(inference_state, consolidated_out["pred_masks_video_res"])
+        return frame_idx, obj_ids, video_res_masks
+
+    def _get_orig_video_res_output(self, inference_state, any_res_masks):
+        """
+        Resize the object scores to the original video resolution (video_res_masks)
+        and apply non-overlapping constraints for final output.
+        """
+        device = inference_state["device"]
+        video_H = inference_state["video_height"]
+        video_W = inference_state["video_width"]
+        any_res_masks = any_res_masks.to(device, non_blocking=True)
+        if any_res_masks.shape[-2:] == (video_H, video_W):
+            video_res_masks = any_res_masks
+        else:
+            video_res_masks = torch.nn.functional.interpolate(
+                any_res_masks,
+                size=(video_H, video_W),
+                mode="bilinear",
+                align_corners=False,
+            )
+        if self.non_overlap_masks:
+            video_res_masks = self._apply_non_overlapping_constraints(video_res_masks)
+        return any_res_masks, video_res_masks
+
+    def _consolidate_temp_output_across_obj(
+        self,
+        inference_state,
+        frame_idx,
+        is_cond,
+        consolidate_at_video_res=False,
+    ):
+        """
+        Consolidate the per-object temporary outputs in `temp_output_dict_per_obj` on
+        a frame into a single output for all objects, including
+        1) fill any missing objects either from `output_dict_per_obj` (if they exist in
+           `output_dict_per_obj` for this frame) or leave them as placeholder values
+           (if they don't exist in `output_dict_per_obj` for this frame);
+        2) if specified, rerun memory encoder after apply non-overlapping constraints
+           on the object scores.
+        """
+        batch_size = self._get_obj_num(inference_state)
+        storage_key = "cond_frame_outputs" if is_cond else "non_cond_frame_outputs"
+        # Optionally, we allow consolidating the temporary outputs at the original
+        # video resolution (to provide a better editing experience for mask prompts).
+        if consolidate_at_video_res:
+            consolidated_H = inference_state["video_height"]
+            consolidated_W = inference_state["video_width"]
+            consolidated_mask_key = "pred_masks_video_res"
+        else:
+            consolidated_H = consolidated_W = self.image_size // 4
+            consolidated_mask_key = "pred_masks"
+
+        # Initialize `consolidated_out`. Its "maskmem_features" and "maskmem_pos_enc"
+        # will be added when rerunning the memory encoder after applying non-overlapping
+        # constraints to object scores. Its "pred_masks" are prefilled with a large
+        # negative value (NO_OBJ_SCORE) to represent missing objects.
+        consolidated_out = {
+            consolidated_mask_key:
+            torch.full(
+                size=(batch_size, 1, consolidated_H, consolidated_W),
+                fill_value=NO_OBJ_SCORE,
+                dtype=inference_state["cached_features"][frame_idx][0].dtype,
+                device=inference_state["storage_device"],
+            ),
+        }
+        for obj_idx in range(batch_size):
+            obj_temp_output_dict = inference_state["temp_output_dict_per_obj"][obj_idx]
+            obj_output_dict = inference_state["output_dict_per_obj"][obj_idx]
+            out = obj_temp_output_dict[storage_key].get(frame_idx, None)
+            # If the object doesn't appear in "temp_output_dict_per_obj" on this frame,
+            # we fall back and look up its previous output in "output_dict_per_obj".
+            # We look up both "cond_frame_outputs" and "non_cond_frame_outputs" in
+            # "output_dict_per_obj" to find a previous output for this object.
+            if out is None:
+                out = obj_output_dict["cond_frame_outputs"].get(frame_idx, None)
+            if out is None:
+                out = obj_output_dict["non_cond_frame_outputs"].get(frame_idx, None)
+            # If the object doesn't appear in "output_dict_per_obj" either, we skip it
+            # and leave its mask scores to the default scores (i.e. the NO_OBJ_SCORE
+            # placeholder above) and set its object pointer to be a dummy pointer.
+            if out is None:
+                continue
+            # Add the temporary object output mask to consolidated output mask
+            obj_mask = out["pred_masks"]
+            consolidated_pred_masks = consolidated_out[consolidated_mask_key]
+            if obj_mask.shape[-2:] == consolidated_pred_masks.shape[-2:]:
+                consolidated_pred_masks[obj_idx:obj_idx + 1] = obj_mask
+            else:
+                # Resize first if temporary object mask has a different resolution
+                resized_obj_mask = torch.nn.functional.interpolate(
+                    obj_mask,
+                    size=consolidated_pred_masks.shape[-2:],
+                    mode="bilinear",
+                    align_corners=False,
+                )
+                consolidated_pred_masks[obj_idx:obj_idx + 1] = resized_obj_mask
+
+        return consolidated_out
+
+    @torch.inference_mode()
+    def propagate_in_video_preflight(self, inference_state):
+        """Prepare inference_state and consolidate temporary outputs before tracking."""
+        # Check and make sure that every object has received input points or masks.
+        batch_size = self._get_obj_num(inference_state)
+        if batch_size == 0:
+            raise RuntimeError("No input points or masks are provided for any object; please add inputs first.")
+
+        # Consolidate per-object temporary outputs in "temp_output_dict_per_obj" and
+        # add them into "output_dict".
+        for obj_idx in range(batch_size):
+            obj_output_dict = inference_state["output_dict_per_obj"][obj_idx]
+            obj_temp_output_dict = inference_state["temp_output_dict_per_obj"][obj_idx]
+            for is_cond in [False, True]:
+                # Separately consolidate conditioning and non-conditioning temp outputs
+                storage_key = ("cond_frame_outputs" if is_cond else "non_cond_frame_outputs")
+                # Find all the frames that contain temporary outputs for any objects
+                # (these should be the frames that have just received clicks for mask inputs
+                # via `add_new_points_or_box` or `add_new_mask`)
+                for frame_idx, out in obj_temp_output_dict[storage_key].items():
+                    # Run memory encoder on the temporary outputs (if the memory feature is missing)
+                    if out["maskmem_features"] is None:
+                        high_res_masks = torch.nn.functional.interpolate(
+                            out["pred_masks"].to(inference_state["device"]),
+                            size=(self.image_size, self.image_size),
+                            mode="bilinear",
+                            align_corners=False,
+                        )
+                        maskmem_features, maskmem_pos_enc = self._run_memory_encoder(
+                            inference_state=inference_state,
+                            frame_idx=frame_idx,
+                            batch_size=1,  # run on the slice of a single object
+                            high_res_masks=high_res_masks,
+                            object_score_logits=out["object_score_logits"],
+                            # these frames are what the user interacted with
+                            is_mask_from_pts=True,
+                        )
+                        out["maskmem_features"] = maskmem_features
+                        out["maskmem_pos_enc"] = maskmem_pos_enc
+
+                    obj_output_dict[storage_key][frame_idx] = out
+                    if self.clear_non_cond_mem_around_input:
+                        # clear non-conditioning memory of the surrounding frames
+                        self._clear_obj_non_cond_mem_around_input(inference_state, frame_idx, obj_idx)
+
+                # clear temporary outputs in `temp_output_dict_per_obj`
+                obj_temp_output_dict[storage_key].clear()
+
+            # check and make sure that every object has received input points or masks
+            obj_output_dict = inference_state["output_dict_per_obj"][obj_idx]
+            if len(obj_output_dict["cond_frame_outputs"]) == 0:
+                obj_id = self._obj_idx_to_id(inference_state, obj_idx)
+                raise RuntimeError(
+                    f"No input points or masks are provided for object id {obj_id}; please add inputs first.")
+            # edge case: if an output is added to "cond_frame_outputs", we remove any prior
+            # output on the same frame in "non_cond_frame_outputs"
+            for frame_idx in obj_output_dict["cond_frame_outputs"]:
+                obj_output_dict["non_cond_frame_outputs"].pop(frame_idx, None)
+
+    @torch.inference_mode()
+    def propagate_in_video(
+        self,
+        inference_state,
+        start_frame_idx=None,
+        max_frame_num_to_track=None,
+        reverse=False,
+        verbose=True,
+    ):
+        """Propagate the input points across frames to track in the entire video."""
+        self.propagate_in_video_preflight(inference_state)
+
+        obj_ids = inference_state["obj_ids"]
+        num_frames = inference_state["num_frames"]
+        batch_size = self._get_obj_num(inference_state)
+
+        # set start index, end index, and processing order
+        if start_frame_idx is None:
+            # default: start from the earliest frame with input points
+            start_frame_idx = min(t for obj_output_dict in inference_state["output_dict_per_obj"].values()
+                                  for t in obj_output_dict["cond_frame_outputs"])
+        if max_frame_num_to_track is None:
+            # default: track all the frames in the video
+            max_frame_num_to_track = num_frames
+        if reverse:
+            end_frame_idx = max(start_frame_idx - max_frame_num_to_track, 0)
+            if start_frame_idx > 0:
+                processing_order = range(start_frame_idx, end_frame_idx - 1, -1)
+            else:
+                processing_order = []  # skip reverse tracking if starting from frame 0
+        else:
+            end_frame_idx = min(start_frame_idx + max_frame_num_to_track, num_frames - 1)
+            processing_order = range(start_frame_idx, end_frame_idx + 1)
+
+        for frame_idx in tqdm(processing_order, desc="propagate in video", disable=not verbose):
+            pred_masks_per_obj = [None] * batch_size
+            for obj_idx in range(batch_size):
+                obj_output_dict = inference_state["output_dict_per_obj"][obj_idx]
+                # We skip those frames already in consolidated outputs (these are frames
+                # that received input clicks or mask). Note that we cannot directly run
+                # batched forward on them via `_run_single_frame_inference` because the
+                # number of clicks on each object might be different.
+                if frame_idx in obj_output_dict["cond_frame_outputs"]:
+                    storage_key = "cond_frame_outputs"
+                    current_out = obj_output_dict[storage_key][frame_idx]
+                    device = inference_state["device"]
+                    pred_masks = current_out["pred_masks"].to(device, non_blocking=True)
+                    if self.clear_non_cond_mem_around_input:
+                        # clear non-conditioning memory of the surrounding frames
+                        self._clear_obj_non_cond_mem_around_input(inference_state, frame_idx, obj_idx)
+                else:
+                    storage_key = "non_cond_frame_outputs"
+                    current_out, pred_masks = self._run_single_frame_inference(
+                        inference_state=inference_state,
+                        output_dict=obj_output_dict,
+                        frame_idx=frame_idx,
+                        batch_size=1,  # run on the slice of a single object
+                        is_init_cond_frame=False,
+                        point_inputs=None,
+                        mask_inputs=None,
+                        hidden_inputs=None,
+                        reverse=reverse,
+                        run_mem_encoder=True,
+                    )
+                    obj_output_dict[storage_key][frame_idx] = current_out
+
+                inference_state["frames_tracked_per_obj"][obj_idx][frame_idx] = {"reverse": reverse}
+                pred_masks_per_obj[obj_idx] = pred_masks
+
+            # Resize the output mask to the original video resolution (we directly use
+            # the mask scores on GPU for output to avoid any CPU conversion in between)
+            if len(pred_masks_per_obj) > 1:
+                all_pred_masks = torch.cat(pred_masks_per_obj, dim=0)
+            else:
+                all_pred_masks = pred_masks_per_obj[0]
+            _, video_res_masks = self._get_orig_video_res_output(inference_state, all_pred_masks)
+            yield frame_idx, obj_ids, video_res_masks
+
+    @torch.inference_mode()
+    def clear_all_prompts_in_frame(self, inference_state, frame_idx, obj_id, need_output=True):
+        """Remove all input points or mask in a specific frame for a given object."""
+        obj_idx = self._obj_id_to_idx(inference_state, obj_id)
+
+        # Clear the conditioning information on the given frame
+        inference_state["point_inputs_per_obj"][obj_idx].pop(frame_idx, None)
+        inference_state["mask_inputs_per_obj"][obj_idx].pop(frame_idx, None)
+
+        temp_output_dict_per_obj = inference_state["temp_output_dict_per_obj"]
+        temp_output_dict_per_obj[obj_idx]["cond_frame_outputs"].pop(frame_idx, None)
+        temp_output_dict_per_obj[obj_idx]["non_cond_frame_outputs"].pop(frame_idx, None)
+
+        # Remove the frame's conditioning output (possibly downgrading it to non-conditioning)
+        obj_output_dict = inference_state["output_dict_per_obj"][obj_idx]
+        out = obj_output_dict["cond_frame_outputs"].pop(frame_idx, None)
+        if out is not None:
+            # The frame is not a conditioning frame anymore since it's not receiving inputs,
+            # so we "downgrade" its output (if exists) to a non-conditioning frame output.
+            obj_output_dict["non_cond_frame_outputs"][frame_idx] = out
+            inference_state["frames_tracked_per_obj"][obj_idx].pop(frame_idx, None)
+
+        if not need_output:
+            return
+        # Finally, output updated masks per object (after removing the inputs above)
+        obj_ids = inference_state["obj_ids"]
+        is_cond = any(frame_idx in obj_temp_output_dict["cond_frame_outputs"]
+                      for obj_temp_output_dict in temp_output_dict_per_obj.values())
+        consolidated_out = self._consolidate_temp_output_across_obj(
+            inference_state,
+            frame_idx,
+            is_cond=is_cond,
+            consolidate_at_video_res=True,
+        )
+        _, video_res_masks = self._get_orig_video_res_output(inference_state, consolidated_out["pred_masks_video_res"])
+        return frame_idx, obj_ids, video_res_masks
+
+    @torch.inference_mode()
+    def reset_state(self, inference_state):
+        """Remove all input points or mask in all frames throughout the video."""
+        self._reset_tracking_results(inference_state)
+        # Remove all object ids
+        inference_state["obj_id_to_idx"].clear()
+        inference_state["obj_idx_to_id"].clear()
+        inference_state["obj_ids"].clear()
+        inference_state["point_inputs_per_obj"].clear()
+        inference_state["mask_inputs_per_obj"].clear()
+        inference_state["output_dict_per_obj"].clear()
+        inference_state["temp_output_dict_per_obj"].clear()
+        inference_state["frames_tracked_per_obj"].clear()
+
+    def _reset_tracking_results(self, inference_state):
+        """Reset all tracking inputs and results across the videos."""
+        for v in inference_state["point_inputs_per_obj"].values():
+            v.clear()
+        for v in inference_state["mask_inputs_per_obj"].values():
+            v.clear()
+        for v in inference_state["output_dict_per_obj"].values():
+            v["cond_frame_outputs"].clear()
+            v["non_cond_frame_outputs"].clear()
+        for v in inference_state["temp_output_dict_per_obj"].values():
+            v["cond_frame_outputs"].clear()
+            v["non_cond_frame_outputs"].clear()
+        for v in inference_state["frames_tracked_per_obj"].values():
+            v.clear()
+
+    def _get_image_feature(self, inference_state, frame_idx, batch_size):
+        """Compute the image features on a given frame."""
+        # NOTE: check me ======================================================================
+        # # Look up in the cache first
+        # image, backbone_out = inference_state["cached_features"].get(frame_idx, (None, None))
+        # if backbone_out is None:
+        #     # Cache miss -- we will run inference on a single image
+        #     device = inference_state["device"]
+        #     image = inference_state["images"][frame_idx].to(device).unsqueeze(0)
+        #     backbone_out = self.forward_image(image)
+        #     # Cache the most recent frame's feature (for repeated interactions with
+        #     # a frame; we can use an LRU cache for more frames in the future).
+        #     inference_state["cached_features"] = {frame_idx: (image, backbone_out)}
+        # =====================================================================================
+
+        # build cache for image features
+        if not inference_state["cached_features"]:
+            image = inference_state["images"].to(inference_state["device"])
+            backbone_out = self.forward_image(image)
+            inference_state["cached_features"] = {
+                i: (image[i, None], {
+                    k: v[i, None] if torch.is_tensor(v) else [t[i, None] for t in v]
+                    for k, v in backbone_out.items()
+                })
+                for i in range(image.size(0))
+            }
+
+        # retrieve from cache
+        image, backbone_out = inference_state["cached_features"][frame_idx]
+
+        # expand the features to have the same dimension as the number of objects
+        expanded_image = image.expand(batch_size, -1, -1, -1)
+        expanded_backbone_out = {
+            "backbone_fpn": backbone_out["backbone_fpn"].copy(),
+            "vision_pos_enc": backbone_out["vision_pos_enc"].copy(),
+        }
+        for i, feat in enumerate(expanded_backbone_out["backbone_fpn"]):
+            expanded_backbone_out["backbone_fpn"][i] = feat.expand(batch_size, -1, -1, -1)
+        for i, pos in enumerate(expanded_backbone_out["vision_pos_enc"]):
+            pos = pos.expand(batch_size, -1, -1, -1)
+            expanded_backbone_out["vision_pos_enc"][i] = pos
+
+        features = self._prepare_backbone_features(expanded_backbone_out)
+        features = (expanded_image, ) + features
+        return features
+
+    def _run_single_frame_inference(
+        self,
+        inference_state,
+        output_dict,
+        frame_idx,
+        batch_size,
+        is_init_cond_frame,
+        point_inputs,
+        mask_inputs,
+        hidden_inputs,
+        reverse,
+        run_mem_encoder,
+        prev_sam_mask_logits=None,
+    ):
+        """Run tracking on a single frame based on current inputs and previous memory."""
+        # Retrieve correct image features
+        (
+            _,
+            _,
+            current_vision_feats,
+            current_vision_pos_embeds,
+            feat_sizes,
+        ) = self._get_image_feature(inference_state, frame_idx, batch_size)
+
+        # point and mask should not appear as input simultaneously on the same frame
+        assert point_inputs is None or mask_inputs is None
+        current_out = self.track_step(
+            frame_idx=frame_idx,
+            is_init_cond_frame=is_init_cond_frame,
+            current_vision_feats=current_vision_feats,
+            current_vision_pos_embeds=current_vision_pos_embeds,
+            feat_sizes=feat_sizes,
+            point_inputs=point_inputs,
+            mask_inputs=mask_inputs,
+            hidden_inputs=hidden_inputs,
+            output_dict=output_dict,
+            num_frames=inference_state["num_frames"],
+            track_in_reverse=reverse,
+            run_mem_encoder=run_mem_encoder,
+            prev_sam_mask_logits=prev_sam_mask_logits,
+        )
+
+        # optionally offload the output to CPU memory to save GPU space
+        storage_device = inference_state["storage_device"]
+        maskmem_features = current_out["maskmem_features"]
+        if maskmem_features is not None:
+            maskmem_features = maskmem_features.to(inference_state["cached_features"][frame_idx][0].dtype)
+            maskmem_features = maskmem_features.to(storage_device, non_blocking=True)
+        pred_masks_gpu = current_out["pred_masks"]
+        # potentially fill holes in the predicted masks
+        if self.fill_hole_area > 0:
+            pred_masks_gpu = fill_holes_in_mask_scores(pred_masks_gpu, self.fill_hole_area)
+        pred_masks = pred_masks_gpu.to(storage_device, non_blocking=True)
+        # "maskmem_pos_enc" is the same across frames, so we only need to store one copy of it
+        maskmem_pos_enc = self._get_maskmem_pos_enc(inference_state, current_out)
+        # object pointer is a small tensor, so we always keep it on GPU memory for fast access
+        obj_ptr = current_out["obj_ptr"]
+        object_score_logits = current_out["object_score_logits"]
+        # make a compact version of this frame's output to reduce the state size
+        compact_current_out = {
+            "maskmem_features": maskmem_features,
+            "maskmem_pos_enc": maskmem_pos_enc,
+            "pred_masks": pred_masks,
+            "obj_ptr": obj_ptr,
+            "object_score_logits": object_score_logits,
+        }
+        # NOTE: reduce memory during inference ----------------------------------------
+        # https://github.com/facebookresearch/sam2/issues/196
+        # step = self.num_maskmem * self.memory_temporal_stride_for_eval * 2
+        # drop_frame_inds = [
+        #     i for i in output_dict["non_cond_frame_outputs"].keys()
+        #     if (i > frame_idx + step if reverse else i < frame_idx - step)
+        # ]
+        # for idx in drop_frame_inds:
+        #     output_dict["non_cond_frame_outputs"].pop(idx)
+        #     for obj_output_dict in inference_state["output_dict_per_obj"].values():
+        #         obj_output_dict["non_cond_frame_outputs"].pop(idx, None)
+        # -----------------------------------------------------------------------------
+        return compact_current_out, pred_masks_gpu
+
+    def _run_memory_encoder(
+        self,
+        inference_state,
+        frame_idx,
+        batch_size,
+        high_res_masks,
+        object_score_logits,
+        is_mask_from_pts,
+    ):
+        """
+        Run the memory encoder on `high_res_masks`. This is usually after applying
+        non-overlapping constraints to object scores. Since their scores changed, their
+        memory also need to be computed again with the memory encoder.
+        """
+        # Retrieve correct image features
+        _, _, current_vision_feats, _, feat_sizes = self._get_image_feature(inference_state, frame_idx, batch_size)
+        maskmem_features, maskmem_pos_enc = self._encode_new_memory(
+            current_vision_feats=current_vision_feats,
+            feat_sizes=feat_sizes,
+            pred_masks_high_res=high_res_masks,
+            object_score_logits=object_score_logits,
+            is_mask_from_pts=is_mask_from_pts,
+        )
+
+        # optionally offload the output to CPU memory to save GPU space
+        storage_device = inference_state["storage_device"]
+        maskmem_features = maskmem_features.to(inference_state["cached_features"][frame_idx][0].dtype)
+        maskmem_features = maskmem_features.to(storage_device, non_blocking=True)
+        # "maskmem_pos_enc" is the same across frames, so we only need to store one copy of it
+        maskmem_pos_enc = self._get_maskmem_pos_enc(inference_state, {"maskmem_pos_enc": maskmem_pos_enc})
+        return maskmem_features, maskmem_pos_enc
+
+    def _get_maskmem_pos_enc(self, inference_state, current_out):
+        """
+        `maskmem_pos_enc` is the same across frames and objects, so we cache it as
+        a constant in the inference session to reduce session storage size.
+        """
+        model_constants = inference_state["constants"]
+        # "out_maskmem_pos_enc" should be either a list of tensors or None
+        out_maskmem_pos_enc = current_out["maskmem_pos_enc"]
+        if out_maskmem_pos_enc is not None:
+            if "maskmem_pos_enc" not in model_constants:
+                assert isinstance(out_maskmem_pos_enc, list)
+                # only take the slice for one object, since it's same across objects
+                maskmem_pos_enc = [x[0:1].clone() for x in out_maskmem_pos_enc]
+                model_constants["maskmem_pos_enc"] = maskmem_pos_enc
+            else:
+                maskmem_pos_enc = model_constants["maskmem_pos_enc"]
+            # expand the cached maskmem_pos_enc to the actual batch size
+            batch_size = out_maskmem_pos_enc[0].size(0)
+            expanded_maskmem_pos_enc = [x.expand(batch_size, -1, -1, -1) for x in maskmem_pos_enc]
+        else:
+            expanded_maskmem_pos_enc = None
+        return expanded_maskmem_pos_enc
+
+    @torch.inference_mode()
+    def remove_object(self, inference_state, obj_id, strict=False, need_output=True):
+        """
+        Remove an object id from the tracking state. If strict is True, we check whether
+        the object id actually exists and raise an error if it doesn't exist.
+        """
+        old_obj_idx_to_rm = inference_state["obj_id_to_idx"].get(obj_id, None)
+        updated_frames = []
+        # Check whether this object_id to remove actually exists and possibly raise an error.
+        if old_obj_idx_to_rm is None:
+            if not strict:
+                return inference_state["obj_ids"], updated_frames
+            raise RuntimeError(f"Cannot remove object id {obj_id} as it doesn't exist. "
+                               f"All existing object ids: {inference_state['obj_ids']}.")
+
+        # If this is the only remaining object id, we simply reset the state.
+        if len(inference_state["obj_id_to_idx"]) == 1:
+            self.reset_state(inference_state)
+            return inference_state["obj_ids"], updated_frames
+
+        # There are still remaining objects after removing this object id. In this case,
+        # we need to delete the object storage from inference state tensors.
+        # Step 0: clear the input on those frames where this object id has point or mask input
+        # (note that this step is required as it might downgrade conditioning frames to
+        # non-conditioning ones)
+        obj_input_frames_inds = set()
+        obj_input_frames_inds.update(inference_state["point_inputs_per_obj"][old_obj_idx_to_rm])
+        obj_input_frames_inds.update(inference_state["mask_inputs_per_obj"][old_obj_idx_to_rm])
+        for frame_idx in obj_input_frames_inds:
+            self.clear_all_prompts_in_frame(inference_state, frame_idx, obj_id, need_output=False)
+
+        # Step 1: Update the object id mapping (note that it must be done after Step 0,
+        # since Step 0 still requires the old object id mappings in inference_state)
+        old_obj_ids = inference_state["obj_ids"]
+        old_obj_inds = list(range(len(old_obj_ids)))
+        remain_old_obj_inds = old_obj_inds.copy()
+        remain_old_obj_inds.remove(old_obj_idx_to_rm)
+        new_obj_ids = [old_obj_ids[old_idx] for old_idx in remain_old_obj_inds]
+        new_obj_inds = list(range(len(new_obj_ids)))
+        # build new mappings
+        old_idx_to_new_idx = dict(zip(remain_old_obj_inds, new_obj_inds))
+        inference_state["obj_id_to_idx"] = dict(zip(new_obj_ids, new_obj_inds))
+        inference_state["obj_idx_to_id"] = dict(zip(new_obj_inds, new_obj_ids))
+        inference_state["obj_ids"] = new_obj_ids
+
+        # Step 2: For per-object tensor storage, we shift their obj_idx in the dict keys.
+        def _map_keys(container):
+            new_kvs = []
+            for k in old_obj_inds:
+                v = container.pop(k)
+                if k in old_idx_to_new_idx:
+                    new_kvs.append((old_idx_to_new_idx[k], v))
+            container.update(new_kvs)
+
+        _map_keys(inference_state["point_inputs_per_obj"])
+        _map_keys(inference_state["mask_inputs_per_obj"])
+        _map_keys(inference_state["output_dict_per_obj"])
+        _map_keys(inference_state["temp_output_dict_per_obj"])
+        _map_keys(inference_state["frames_tracked_per_obj"])
+
+        # Step 3: Further collect the outputs on those frames in `obj_input_frames_inds`, which
+        # could show an updated mask for objects previously occluded by the object being removed
+        if need_output:
+            temp_output_dict_per_obj = inference_state["temp_output_dict_per_obj"]
+            for frame_idx in obj_input_frames_inds:
+                is_cond = any(frame_idx in obj_temp_output_dict["cond_frame_outputs"]
+                              for obj_temp_output_dict in temp_output_dict_per_obj.values())
+                consolidated_out = self._consolidate_temp_output_across_obj(
+                    inference_state,
+                    frame_idx,
+                    is_cond=is_cond,
+                    consolidate_at_video_res=True,
+                )
+                _, video_res_masks = self._get_orig_video_res_output(inference_state,
+                                                                     consolidated_out["pred_masks_video_res"])
+                updated_frames.append((frame_idx, video_res_masks))
+
+        return inference_state["obj_ids"], updated_frames
+
+    def _clear_non_cond_mem_around_input(self, inference_state, frame_idx):
+        """
+        Remove the non-conditioning memory around the input frame. When users provide
+        correction clicks, the surrounding frames' non-conditioning memories can still
+        contain outdated object appearance information and could confuse the model.
+
+        This method clears those non-conditioning memories surrounding the interacted
+        frame to avoid giving the model both old and new information about the object.
+        """
+        r = self.memory_temporal_stride_for_eval
+        frame_idx_begin = frame_idx - r * self.num_maskmem
+        frame_idx_end = frame_idx + r * self.num_maskmem
+        batch_size = self._get_obj_num(inference_state)
+        for obj_idx in range(batch_size):
+            obj_output_dict = inference_state["output_dict_per_obj"][obj_idx]
+            non_cond_frame_outputs = obj_output_dict["non_cond_frame_outputs"]
+            for t in range(frame_idx_begin, frame_idx_end + 1):
+                non_cond_frame_outputs.pop(t, None)
+
+
+class SAM2VideoPredictorVOS(SAM2VideoPredictor):
+    """Optimized for the VOS setting"""
+
+    def __init__(self, *args, **kwargs):
+        raise NotImplementedError("SAM2VideoPredictorVOS has not been modified for LLMs")
+        super().__init__(*args, **kwargs)
+        self._compile_all_components()
+
+    def _compile_all_components(self):
+        print("Compiling all components for VOS setting. First time may be very slow.")
+        self.memory_encoder.forward = torch.compile(
+            self.memory_encoder.forward,
+            mode="max-autotune",
+            fullgraph=True,
+            dynamic=False,
+        )
+
+        self.memory_attention.forward = torch.compile(
+            self.memory_attention.forward,
+            mode="max-autotune",
+            fullgraph=True,
+            dynamic=True,  # Num. of memories varies
+        )
+
+        self.sam_prompt_encoder.forward = torch.compile(
+            self.sam_prompt_encoder.forward,
+            mode="max-autotune",
+            fullgraph=True,
+            dynamic=False,  # Accuracy regression on True
+        )
+
+        self.sam_mask_decoder.forward = torch.compile(
+            self.sam_mask_decoder.forward,
+            mode="max-autotune",
+            fullgraph=True,
+            dynamic=False,  # Accuracy regression on True
+        )
+
+    def forward_image(self, img_batch: torch.Tensor):
+        """
+        Identical to the corresponding method in the parent (SAM2VideoPredictor), but
+        cloning the backbone features and pos encoding to enable compilation.
+        """
+        backbone_out = self.image_encoder(img_batch)
+        if self.use_high_res_features_in_sam:
+            # precompute projected level 0 and level 1 features in SAM decoder
+            # to avoid running it again on every SAM click
+            backbone_out["backbone_fpn"][0] = self.sam_mask_decoder.conv_s0(backbone_out["backbone_fpn"][0])
+            backbone_out["backbone_fpn"][1] = self.sam_mask_decoder.conv_s1(backbone_out["backbone_fpn"][1])
+        # Clone to help torch.compile
+        for i in range(len(backbone_out["backbone_fpn"])):
+            backbone_out["backbone_fpn"][i] = backbone_out["backbone_fpn"][i].clone()
+            backbone_out["vision_pos_enc"][i] = backbone_out["vision_pos_enc"][i].clone()
+        return backbone_out
+
+    def _forward_sam_heads(
+        self,
+        backbone_features,
+        point_inputs=None,
+        mask_inputs=None,
+        high_res_features=None,
+        multimask_output=False,
+    ):
+        """
+        Identical to the corresponding method in the parent (SAM2VideoPredictor), but
+        cloning the outputs of prompt_encoder and mask_decoder to enable compilation.
+        """
+        B = backbone_features.size(0)
+        device = backbone_features.device
+        assert backbone_features.size(1) == self.sam_prompt_embed_dim
+        assert backbone_features.size(2) == self.sam_image_embedding_size
+        assert backbone_features.size(3) == self.sam_image_embedding_size
+
+        # a) Handle point prompts
+        if point_inputs is not None:
+            sam_point_coords = point_inputs["point_coords"]
+            sam_point_labels = point_inputs["point_labels"]
+            assert sam_point_coords.size(0) == B and sam_point_labels.size(0) == B
+        else:
+            # If no points are provide, pad with an empty point (with label -1)
+            sam_point_coords = torch.zeros(B, 1, 2, device=device)
+            sam_point_labels = -torch.ones(B, 1, dtype=torch.int32, device=device)
+
+        # b) Handle mask prompts
+        if mask_inputs is not None:
+            # If mask_inputs is provided, downsize it into low-res mask input if needed
+            # and feed it as a dense mask prompt into the SAM mask encoder
+            assert len(mask_inputs.shape) == 4 and mask_inputs.shape[:2] == (B, 1)
+            if mask_inputs.shape[-2:] != self.sam_prompt_encoder.mask_input_size:
+                sam_mask_prompt = F.interpolate(
+                    mask_inputs.float(),
+                    size=self.sam_prompt_encoder.mask_input_size,
+                    align_corners=False,
+                    mode="bilinear",
+                    antialias=True,  # use antialias for downsampling
+                )
+            else:
+                sam_mask_prompt = mask_inputs
+        else:
+            # Otherwise, simply feed None (and SAM's prompt encoder will add
+            # a learned `no_mask_embed` to indicate no mask input in this case).
+            sam_mask_prompt = None
+
+        sparse_embeddings, dense_embeddings = self.sam_prompt_encoder(
+            points=(sam_point_coords, sam_point_labels),
+            boxes=None,
+            masks=sam_mask_prompt,
+        )
+        # Clone image_pe and the outputs of sam_prompt_encoder
+        # to enable compilation
+        sparse_embeddings = sparse_embeddings.clone()
+        dense_embeddings = dense_embeddings.clone()
+        image_pe = self.sam_prompt_encoder.get_dense_pe().clone()
+        (
+            low_res_multimasks,
+            ious,
+            sam_output_tokens,
+            object_score_logits,
+        ) = self.sam_mask_decoder(
+            image_embeddings=backbone_features,
+            image_pe=image_pe,
+            sparse_prompt_embeddings=sparse_embeddings,
+            dense_prompt_embeddings=dense_embeddings,
+            multimask_output=multimask_output,
+            repeat_image=False,  # the image is already batched
+            high_res_features=high_res_features,
+        )
+        # Clone the output of sam_mask_decoder
+        # to enable compilation
+        low_res_multimasks = low_res_multimasks.clone()
+        ious = ious.clone()
+        sam_output_tokens = sam_output_tokens.clone()
+        object_score_logits = object_score_logits.clone()
+
+        if self.pred_obj_scores:
+            is_obj_appearing = object_score_logits > 0
+
+            # Mask used for spatial memories is always a *hard* choice between obj and no obj,
+            # consistent with the actual mask prediction
+            low_res_multimasks = torch.where(
+                is_obj_appearing[:, None, None],
+                low_res_multimasks,
+                NO_OBJ_SCORE,
+            )
+
+        # convert masks from possibly bfloat16 (or float16) to float32
+        low_res_multimasks = low_res_multimasks.float()
+        high_res_multimasks = F.interpolate(
+            low_res_multimasks,
+            size=(self.image_size, self.image_size),
+            mode="bilinear",
+            align_corners=False,
+        )
+
+        sam_output_token = sam_output_tokens[:, 0]
+        if multimask_output:
+            # take the best mask prediction (with the highest IoU estimation)
+            best_iou_inds = torch.argmax(ious, dim=-1)
+            batch_inds = torch.arange(B, device=device)
+            low_res_masks = low_res_multimasks[batch_inds, best_iou_inds].unsqueeze(1)
+            high_res_masks = high_res_multimasks[batch_inds, best_iou_inds].unsqueeze(1)
+            if sam_output_tokens.size(1) > 1:
+                sam_output_token = sam_output_tokens[batch_inds, best_iou_inds]
+        else:
+            low_res_masks, high_res_masks = low_res_multimasks, high_res_multimasks
+
+        # Extract object pointer from the SAM output token (with occlusion handling)
+        obj_ptr = self.obj_ptr_proj(sam_output_token)
+        if self.pred_obj_scores:
+            # Allow *soft* no obj ptr, unlike for masks
+            if self.soft_no_obj_ptr:
+                lambda_is_obj_appearing = object_score_logits.sigmoid()
+            else:
+                lambda_is_obj_appearing = is_obj_appearing.float()
+
+            if self.fixed_no_obj_ptr:
+                obj_ptr = lambda_is_obj_appearing * obj_ptr
+            obj_ptr = obj_ptr + (1 - lambda_is_obj_appearing) * self.no_obj_ptr
+
+        return (
+            low_res_multimasks,
+            high_res_multimasks,
+            ious,
+            low_res_masks,
+            high_res_masks,
+            obj_ptr,
+            object_score_logits,
+        )
+
+    def _encode_new_memory(
+        self,
+        current_vision_feats,
+        feat_sizes,
+        pred_masks_high_res,
+        object_score_logits,
+        is_mask_from_pts,
+    ):
+        """
+        Identical to the corresponding method in the parent (SAM2VideoPredictor), but
+        cloning the memories and their pos enc to enable compilation.
+        """
+        B = current_vision_feats[-1].size(1)  # batch size on this frame
+        C = self.hidden_dim
+        H, W = feat_sizes[-1]  # top-level (lowest-resolution) feature size
+        # top-level feature, (HW)BC => BCHW
+        pix_feat = current_vision_feats[-1].permute(1, 2, 0).view(B, C, H, W)
+        if self.non_overlap_masks_for_mem_enc and not self.training:
+            # optionally, apply non-overlapping constraints to the masks (it's applied
+            # in the batch dimension and should only be used during eval, where all
+            # the objects come from the same video under batch size 1).
+            pred_masks_high_res = self._apply_non_overlapping_constraints(pred_masks_high_res)
+        # scale the raw mask logits with a temperature before applying sigmoid
+        binarize = self.binarize_mask_from_pts_for_mem_enc and is_mask_from_pts
+        if binarize and not self.training:
+            mask_for_mem = (pred_masks_high_res > 0).float()
+        else:
+            # apply sigmoid on the raw mask logits to turn them into range (0, 1)
+            mask_for_mem = torch.sigmoid(pred_masks_high_res)
+        # apply scale and bias terms to the sigmoid probabilities
+        if self.sigmoid_scale_for_mem_enc != 1.0:
+            mask_for_mem = mask_for_mem * self.sigmoid_scale_for_mem_enc
+        if self.sigmoid_bias_for_mem_enc != 0.0:
+            mask_for_mem = mask_for_mem + self.sigmoid_bias_for_mem_enc
+        maskmem_out = self.memory_encoder(
+            pix_feat,
+            mask_for_mem,
+            skip_mask_sigmoid=True  # sigmoid already applied
+        )
+        # Clone the feats and pos_enc to enable compilation
+        maskmem_features = maskmem_out["vision_features"].clone()
+        maskmem_pos_enc = [m.clone() for m in maskmem_out["vision_pos_enc"]]
+        # add a no-object embedding to the spatial memory to indicate that the frame
+        # is predicted to be occluded (i.e. no object is appearing in the frame)
+        if self.no_obj_embed_spatial is not None:
+            is_obj_appearing = (object_score_logits > 0).float()
+            maskmem_features += (1 - is_obj_appearing[..., None, None]
+                                 ) * self.no_obj_embed_spatial[..., None, None].expand(*maskmem_features.shape)
+
+        return maskmem_features, maskmem_pos_enc

sam2/utils/__init__.py

@@ -0,0 +1,5 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.

sam2/utils/amg.py

@@ -0,0 +1,328 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.
+
+import math
+from copy import deepcopy
+from itertools import product
+from typing import Any, Dict, Generator, ItemsView, List, Tuple
+
+import numpy as np
+import torch
+
+# Very lightly adapted from https://github.com/facebookresearch/segment-anything/blob/main/segment_anything/utils/amg.py
+
+
+class MaskData:
+    """
+    A structure for storing masks and their related data in batched format.
+    Implements basic filtering and concatenation.
+    """
+
+    def __init__(self, **kwargs) -> None:
+        for v in kwargs.values():
+            assert isinstance(
+                v, (list, np.ndarray, torch.Tensor)), "MaskData only supports list, numpy arrays, and torch tensors."
+        self._stats = dict(**kwargs)
+
+    def __setitem__(self, key: str, item: Any) -> None:
+        assert isinstance(
+            item, (list, np.ndarray, torch.Tensor)), "MaskData only supports list, numpy arrays, and torch tensors."
+        self._stats[key] = item
+
+    def __delitem__(self, key: str) -> None:
+        del self._stats[key]
+
+    def __getitem__(self, key: str) -> Any:
+        return self._stats[key]
+
+    def items(self) -> ItemsView[str, Any]:
+        return self._stats.items()
+
+    def filter(self, keep: torch.Tensor) -> None:
+        for k, v in self._stats.items():
+            if v is None:
+                self._stats[k] = None
+            elif isinstance(v, torch.Tensor):
+                self._stats[k] = v[torch.as_tensor(keep, device=v.device)]
+            elif isinstance(v, np.ndarray):
+                self._stats[k] = v[keep.detach().cpu().numpy()]
+            elif isinstance(v, list) and keep.dtype == torch.bool:
+                self._stats[k] = [a for i, a in enumerate(v) if keep[i]]
+            elif isinstance(v, list):
+                self._stats[k] = [v[i] for i in keep]
+            else:
+                raise TypeError(f"MaskData key {k} has an unsupported type {type(v)}.")
+
+    def cat(self, new_stats: "MaskData") -> None:
+        for k, v in new_stats.items():
+            if k not in self._stats or self._stats[k] is None:
+                self._stats[k] = deepcopy(v)
+            elif isinstance(v, torch.Tensor):
+                self._stats[k] = torch.cat([self._stats[k], v], dim=0)
+            elif isinstance(v, np.ndarray):
+                self._stats[k] = np.concatenate([self._stats[k], v], axis=0)
+            elif isinstance(v, list):
+                self._stats[k] = self._stats[k] + deepcopy(v)
+            else:
+                raise TypeError(f"MaskData key {k} has an unsupported type {type(v)}.")
+
+    def to_numpy(self) -> None:
+        for k, v in self._stats.items():
+            if isinstance(v, torch.Tensor):
+                self._stats[k] = v.float().detach().cpu().numpy()
+
+
+def is_box_near_crop_edge(boxes: torch.Tensor,
+                          crop_box: List[int],
+                          orig_box: List[int],
+                          atol: float = 20.0) -> torch.Tensor:
+    """Filter masks at the edge of a crop, but not at the edge of the original image."""
+    crop_box_torch = torch.as_tensor(crop_box, dtype=torch.float, device=boxes.device)
+    orig_box_torch = torch.as_tensor(orig_box, dtype=torch.float, device=boxes.device)
+    boxes = uncrop_boxes_xyxy(boxes, crop_box).float()
+    near_crop_edge = torch.isclose(boxes, crop_box_torch[None, :], atol=atol, rtol=0)
+    near_image_edge = torch.isclose(boxes, orig_box_torch[None, :], atol=atol, rtol=0)
+    near_crop_edge = torch.logical_and(near_crop_edge, ~near_image_edge)
+    return torch.any(near_crop_edge, dim=1)
+
+
+def box_xyxy_to_xywh(box_xyxy: torch.Tensor) -> torch.Tensor:
+    box_xywh = deepcopy(box_xyxy)
+    box_xywh[2] = box_xywh[2] - box_xywh[0]
+    box_xywh[3] = box_xywh[3] - box_xywh[1]
+    return box_xywh
+
+
+def batch_iterator(batch_size: int, *args) -> Generator[List[Any], None, None]:
+    assert len(args) > 0 and all(len(a) == len(args[0])
+                                 for a in args), "Batched iteration must have inputs of all the same size."
+    n_batches = len(args[0]) // batch_size + int(len(args[0]) % batch_size != 0)
+    for b in range(n_batches):
+        yield [arg[b * batch_size:(b + 1) * batch_size] for arg in args]
+
+
+def mask_to_rle_pytorch(tensor: torch.Tensor) -> List[Dict[str, Any]]:
+    """
+    Encodes masks to an uncompressed RLE, in the format expected by
+    pycoco tools.
+    """
+    # Put in fortran order and flatten h,w
+    b, h, w = tensor.shape
+    tensor = tensor.permute(0, 2, 1).flatten(1)
+
+    # Compute change indices
+    diff = tensor[:, 1:] ^ tensor[:, :-1]
+    change_indices = diff.nonzero()
+
+    # Encode run length
+    out = []
+    for i in range(b):
+        cur_idxs = change_indices[change_indices[:, 0] == i, 1]
+        cur_idxs = torch.cat([
+            torch.tensor([0], dtype=cur_idxs.dtype, device=cur_idxs.device),
+            cur_idxs + 1,
+            torch.tensor([h * w], dtype=cur_idxs.dtype, device=cur_idxs.device),
+        ])
+        btw_idxs = cur_idxs[1:] - cur_idxs[:-1]
+        counts = [] if tensor[i, 0] == 0 else [0]
+        counts.extend(btw_idxs.detach().cpu().tolist())
+        out.append({"size": [h, w], "counts": counts})
+    return out
+
+
+def rle_to_mask(rle: Dict[str, Any]) -> np.ndarray:
+    """Compute a binary mask from an uncompressed RLE."""
+    h, w = rle["size"]
+    mask = np.empty(h * w, dtype=bool)
+    idx = 0
+    parity = False
+    for count in rle["counts"]:
+        mask[idx:idx + count] = parity
+        idx += count
+        parity ^= True
+    mask = mask.reshape(w, h)
+    return mask.transpose()  # Put in C order
+
+
+def area_from_rle(rle: Dict[str, Any]) -> int:
+    return sum(rle["counts"][1::2])
+
+
+def calculate_stability_score(masks: torch.Tensor, mask_threshold: float, threshold_offset: float) -> torch.Tensor:
+    """
+    Computes the stability score for a batch of masks. The stability
+    score is the IoU between the binary masks obtained by thresholding
+    the predicted mask logits at high and low values.
+    """
+    # One mask is always contained inside the other.
+    # Save memory by preventing unnecessary cast to torch.int64
+    intersections = ((masks > (mask_threshold + threshold_offset)).sum(-1,
+                                                                       dtype=torch.int16).sum(-1, dtype=torch.int32))
+    unions = ((masks > (mask_threshold - threshold_offset)).sum(-1, dtype=torch.int16).sum(-1, dtype=torch.int32))
+    return intersections / unions
+
+
+def build_point_grid(n_per_side: int) -> np.ndarray:
+    """Generates a 2D grid of points evenly spaced in [0,1]x[0,1]."""
+    offset = 1 / (2 * n_per_side)
+    points_one_side = np.linspace(offset, 1 - offset, n_per_side)
+    points_x = np.tile(points_one_side[None, :], (n_per_side, 1))
+    points_y = np.tile(points_one_side[:, None], (1, n_per_side))
+    points = np.stack([points_x, points_y], axis=-1).reshape(-1, 2)
+    return points
+
+
+def build_all_layer_point_grids(n_per_side: int, n_layers: int, scale_per_layer: int) -> List[np.ndarray]:
+    """Generates point grids for all crop layers."""
+    points_by_layer = []
+    for i in range(n_layers + 1):
+        n_points = int(n_per_side / (scale_per_layer**i))
+        points_by_layer.append(build_point_grid(n_points))
+    return points_by_layer
+
+
+def generate_crop_boxes(im_size: Tuple[int, ...], n_layers: int,
+                        overlap_ratio: float) -> Tuple[List[List[int]], List[int]]:
+    """
+    Generates a list of crop boxes of different sizes. Each layer
+    has (2**i)**2 boxes for the ith layer.
+    """
+    crop_boxes, layer_idxs = [], []
+    im_h, im_w = im_size
+    short_side = min(im_h, im_w)
+
+    # Original image
+    crop_boxes.append([0, 0, im_w, im_h])
+    layer_idxs.append(0)
+
+    def crop_len(orig_len, n_crops, overlap):
+        return int(math.ceil((overlap * (n_crops - 1) + orig_len) / n_crops))
+
+    for i_layer in range(n_layers):
+        n_crops_per_side = 2**(i_layer + 1)
+        overlap = int(overlap_ratio * short_side * (2 / n_crops_per_side))
+
+        crop_w = crop_len(im_w, n_crops_per_side, overlap)
+        crop_h = crop_len(im_h, n_crops_per_side, overlap)
+
+        crop_box_x0 = [int((crop_w - overlap) * i) for i in range(n_crops_per_side)]
+        crop_box_y0 = [int((crop_h - overlap) * i) for i in range(n_crops_per_side)]
+
+        # Crops in XYWH format
+        for x0, y0 in product(crop_box_x0, crop_box_y0):
+            box = [x0, y0, min(x0 + crop_w, im_w), min(y0 + crop_h, im_h)]
+            crop_boxes.append(box)
+            layer_idxs.append(i_layer + 1)
+
+    return crop_boxes, layer_idxs
+
+
+def uncrop_boxes_xyxy(boxes: torch.Tensor, crop_box: List[int]) -> torch.Tensor:
+    x0, y0, _, _ = crop_box
+    offset = torch.tensor([[x0, y0, x0, y0]], device=boxes.device)
+    # Check if boxes has a channel dimension
+    if len(boxes.shape) == 3:
+        offset = offset.unsqueeze(1)
+    return boxes + offset
+
+
+def uncrop_points(points: torch.Tensor, crop_box: List[int]) -> torch.Tensor:
+    x0, y0, _, _ = crop_box
+    offset = torch.tensor([[x0, y0]], device=points.device)
+    # Check if points has a channel dimension
+    if len(points.shape) == 3:
+        offset = offset.unsqueeze(1)
+    return points + offset
+
+
+def uncrop_masks(masks: torch.Tensor, crop_box: List[int], orig_h: int, orig_w: int) -> torch.Tensor:
+    x0, y0, x1, y1 = crop_box
+    if x0 == 0 and y0 == 0 and x1 == orig_w and y1 == orig_h:
+        return masks
+    # Coordinate transform masks
+    pad_x, pad_y = orig_w - (x1 - x0), orig_h - (y1 - y0)
+    pad = (x0, pad_x - x0, y0, pad_y - y0)
+    return torch.nn.functional.pad(masks, pad, value=0)
+
+
+def remove_small_regions(mask: np.ndarray, area_thresh: float, mode: str) -> Tuple[np.ndarray, bool]:
+    """
+    Removes small disconnected regions and holes in a mask. Returns the
+    mask and an indicator of if the mask has been modified.
+    """
+    import cv2  # type: ignore
+
+    assert mode in ["holes", "islands"]
+    correct_holes = mode == "holes"
+    working_mask = (correct_holes ^ mask).astype(np.uint8)
+    n_labels, regions, stats, _ = cv2.connectedComponentsWithStats(working_mask, 8)
+    sizes = stats[:, -1][1:]  # Row 0 is background label
+    small_regions = [i + 1 for i, s in enumerate(sizes) if s < area_thresh]
+    if len(small_regions) == 0:
+        return mask, False
+    fill_labels = [0] + small_regions
+    if not correct_holes:
+        fill_labels = [i for i in range(n_labels) if i not in fill_labels]
+        # If every region is below threshold, keep largest
+        if len(fill_labels) == 0:
+            fill_labels = [int(np.argmax(sizes)) + 1]
+    mask = np.isin(regions, fill_labels)
+    return mask, True
+
+
+def coco_encode_rle(uncompressed_rle: Dict[str, Any]) -> Dict[str, Any]:
+    from pycocotools import mask as mask_utils  # type: ignore
+
+    h, w = uncompressed_rle["size"]
+    rle = mask_utils.frPyObjects(uncompressed_rle, h, w)
+    rle["counts"] = rle["counts"].decode("utf-8")  # Necessary to serialize with json
+    return rle
+
+
+def batched_mask_to_box(masks: torch.Tensor) -> torch.Tensor:
+    """
+    Calculates boxes in XYXY format around masks. Return [0,0,0,0] for
+    an empty mask. For input shape C1xC2x...xHxW, the output shape is C1xC2x...x4.
+    """
+    # torch.max below raises an error on empty inputs, just skip in this case
+    if torch.numel(masks) == 0:
+        return torch.zeros(*masks.shape[:-2], 4, device=masks.device)
+
+    # Normalize shape to CxHxW
+    shape = masks.shape
+    h, w = shape[-2:]
+    if len(shape) > 2:
+        masks = masks.flatten(0, -3)
+    else:
+        masks = masks.unsqueeze(0)
+
+    # Get top and bottom edges
+    in_height, _ = torch.max(masks, dim=-1)
+    in_height_coords = in_height * torch.arange(h, device=in_height.device)[None, :]
+    bottom_edges, _ = torch.max(in_height_coords, dim=-1)
+    in_height_coords = in_height_coords + h * (~in_height)
+    top_edges, _ = torch.min(in_height_coords, dim=-1)
+
+    # Get left and right edges
+    in_width, _ = torch.max(masks, dim=-2)
+    in_width_coords = in_width * torch.arange(w, device=in_width.device)[None, :]
+    right_edges, _ = torch.max(in_width_coords, dim=-1)
+    in_width_coords = in_width_coords + w * (~in_width)
+    left_edges, _ = torch.min(in_width_coords, dim=-1)
+
+    # If the mask is empty the right edge will be to the left of the left edge.
+    # Replace these boxes with [0, 0, 0, 0]
+    empty_filter = (right_edges < left_edges) | (bottom_edges < top_edges)
+    out = torch.stack([left_edges, top_edges, right_edges, bottom_edges], dim=-1)
+    out = out * (~empty_filter).unsqueeze(-1)
+
+    # Return to original shape
+    if len(shape) > 2:
+        out = out.reshape(*shape[:-2], 4)
+    else:
+        out = out[0]
+
+    return out

sam2/utils/misc.py

@@ -0,0 +1,340 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.
+
+import os
+import warnings
+from threading import Thread
+
+import numpy as np
+import torch
+from PIL import Image
+from tqdm import tqdm
+
+
+def get_sdpa_settings():
+    if torch.cuda.is_available():
+        old_gpu = torch.cuda.get_device_properties(0).major < 7
+        # only use Flash Attention on Ampere (8.0) or newer GPUs
+        use_flash_attn = torch.cuda.get_device_properties(0).major >= 8
+        if not use_flash_attn:
+            warnings.warn(
+                "Flash Attention is disabled as it requires a GPU with Ampere (8.0) CUDA capability.",
+                category=UserWarning,
+                stacklevel=2,
+            )
+        # keep math kernel for PyTorch versions before 2.2 (Flash Attention v2 is only
+        # available on PyTorch 2.2+, while Flash Attention v1 cannot handle all cases)
+        pytorch_version = tuple(int(v) for v in torch.__version__.split(".")[:2])
+        if pytorch_version < (2, 2):
+            warnings.warn(
+                f"You are using PyTorch {torch.__version__} without Flash Attention v2 support. "
+                "Consider upgrading to PyTorch 2.2+ for Flash Attention v2 (which could be faster).",
+                category=UserWarning,
+                stacklevel=2,
+            )
+        math_kernel_on = pytorch_version < (2, 2) or not use_flash_attn
+    else:
+        old_gpu = True
+        use_flash_attn = False
+        math_kernel_on = True
+
+    return old_gpu, use_flash_attn, math_kernel_on
+
+
+def get_connected_components(mask):
+    """
+    Get the connected components (8-connectivity) of binary masks of shape (N, 1, H, W).
+
+    Inputs:
+    - mask: A binary mask tensor of shape (N, 1, H, W), where 1 is foreground and 0 is
+            background.
+
+    Outputs:
+    - labels: A tensor of shape (N, 1, H, W) containing the connected component labels
+              for foreground pixels and 0 for background pixels.
+    - counts: A tensor of shape (N, 1, H, W) containing the area of the connected
+              components for foreground pixels and 0 for background pixels.
+    """
+    from sam2 import _C
+
+    return _C.get_connected_componnets(mask.to(torch.uint8).contiguous())
+
+
+def mask_to_box(masks: torch.Tensor):
+    """
+    compute bounding box given an input mask
+
+    Inputs:
+    - masks: [B, 1, H, W] masks, dtype=torch.Tensor
+
+    Returns:
+    - box_coords: [B, 1, 4], contains (x, y) coordinates of top left and bottom right box corners, dtype=torch.Tensor
+    """
+    B, _, h, w = masks.shape
+    device = masks.device
+    xs = torch.arange(w, device=device, dtype=torch.int32)
+    ys = torch.arange(h, device=device, dtype=torch.int32)
+    grid_xs, grid_ys = torch.meshgrid(xs, ys, indexing="xy")
+    grid_xs = grid_xs[None, None, ...].expand(B, 1, h, w)
+    grid_ys = grid_ys[None, None, ...].expand(B, 1, h, w)
+    min_xs, _ = torch.min(torch.where(masks, grid_xs, w).flatten(-2), dim=-1)
+    max_xs, _ = torch.max(torch.where(masks, grid_xs, -1).flatten(-2), dim=-1)
+    min_ys, _ = torch.min(torch.where(masks, grid_ys, h).flatten(-2), dim=-1)
+    max_ys, _ = torch.max(torch.where(masks, grid_ys, -1).flatten(-2), dim=-1)
+    bbox_coords = torch.stack((min_xs, min_ys, max_xs, max_ys), dim=-1)
+
+    return bbox_coords
+
+
+def _load_img_as_tensor(img_path, image_size):
+    img_pil = Image.open(img_path)
+    img_np = np.array(img_pil.convert("RGB").resize((image_size, image_size)))
+    if img_np.dtype == np.uint8:  # np.uint8 is expected for JPEG images
+        img_np = img_np / 255.0
+    else:
+        raise RuntimeError(f"Unknown image dtype: {img_np.dtype} on {img_path}")
+    img = torch.from_numpy(img_np).permute(2, 0, 1)
+    video_width, video_height = img_pil.size  # the original video size
+    return img, video_height, video_width
+
+
+class AsyncVideoFrameLoader:
+    """
+    A list of video frames to be load asynchronously without blocking session start.
+    """
+
+    def __init__(
+        self,
+        img_paths,
+        image_size,
+        offload_video_to_cpu,
+        img_mean,
+        img_std,
+        compute_device,
+    ):
+        self.img_paths = img_paths
+        self.image_size = image_size
+        self.offload_video_to_cpu = offload_video_to_cpu
+        self.img_mean = img_mean
+        self.img_std = img_std
+        # items in `self.images` will be loaded asynchronously
+        self.images = [None] * len(img_paths)
+        # catch and raise any exceptions in the async loading thread
+        self.exception = None
+        # video_height and video_width be filled when loading the first image
+        self.video_height = None
+        self.video_width = None
+        self.compute_device = compute_device
+
+        # load the first frame to fill video_height and video_width and also
+        # to cache it (since it's most likely where the user will click)
+        self.__getitem__(0)
+
+        # load the rest of frames asynchronously without blocking the session start
+        def _load_frames():
+            try:
+                for n in tqdm(range(len(self.images)), desc="frame loading (JPEG)"):
+                    self.__getitem__(n)
+            except Exception as e:
+                self.exception = e
+
+        self.thread = Thread(target=_load_frames, daemon=True)
+        self.thread.start()
+
+    def __getitem__(self, index):
+        if self.exception is not None:
+            raise RuntimeError("Failure in frame loading thread") from self.exception
+
+        img = self.images[index]
+        if img is not None:
+            return img
+
+        img, video_height, video_width = _load_img_as_tensor(self.img_paths[index], self.image_size)
+        self.video_height = video_height
+        self.video_width = video_width
+        # normalize by mean and std
+        img -= self.img_mean
+        img /= self.img_std
+        if not self.offload_video_to_cpu:
+            img = img.to(self.compute_device, non_blocking=True)
+        self.images[index] = img
+        return img
+
+    def __len__(self):
+        return len(self.images)
+
+
+def load_video_frames(
+        video_path,
+        image_size,
+        offload_video_to_cpu,
+        img_mean=(0.485, 0.456, 0.406),
+        img_std=(0.229, 0.224, 0.225),
+        async_loading_frames=False,
+        compute_device=torch.device("cuda"),
+):
+    """
+    Load the video frames from video_path. The frames are resized to image_size as in
+    the model and are loaded to GPU if offload_video_to_cpu=False. This is used by the demo.
+    """
+    is_bytes = isinstance(video_path, bytes)
+    is_str = isinstance(video_path, str)
+    is_mp4_path = is_str and os.path.splitext(video_path)[-1] in [".mp4", ".MP4"]
+    if is_bytes or is_mp4_path:
+        return load_video_frames_from_video_file(
+            video_path=video_path,
+            image_size=image_size,
+            offload_video_to_cpu=offload_video_to_cpu,
+            img_mean=img_mean,
+            img_std=img_std,
+            compute_device=compute_device,
+        )
+    elif is_str and os.path.isdir(video_path):
+        return load_video_frames_from_jpg_images(
+            video_path=video_path,
+            image_size=image_size,
+            offload_video_to_cpu=offload_video_to_cpu,
+            img_mean=img_mean,
+            img_std=img_std,
+            async_loading_frames=async_loading_frames,
+            compute_device=compute_device,
+        )
+    else:
+        raise NotImplementedError("Only MP4 video and JPEG folder are supported at this moment")
+
+
+def load_video_frames_from_jpg_images(
+        video_path,
+        image_size,
+        offload_video_to_cpu,
+        img_mean=(0.485, 0.456, 0.406),
+        img_std=(0.229, 0.224, 0.225),
+        async_loading_frames=False,
+        compute_device=torch.device("cuda"),
+):
+    """
+    Load the video frames from a directory of JPEG files ("<frame_index>.jpg" format).
+
+    The frames are resized to image_size x image_size and are loaded to GPU if
+    `offload_video_to_cpu` is `False` and to CPU if `offload_video_to_cpu` is `True`.
+
+    You can load a frame asynchronously by setting `async_loading_frames` to `True`.
+    """
+    if isinstance(video_path, str) and os.path.isdir(video_path):
+        jpg_folder = video_path
+    else:
+        raise NotImplementedError(
+            "Only JPEG frames are supported at this moment. For video files, you may use "
+            "ffmpeg (https://ffmpeg.org/) to extract frames into a folder of JPEG files, such as \n"
+            "```\n"
+            "ffmpeg -i <your_video>.mp4 -q:v 2 -start_number 0 <output_dir>/'%05d.jpg'\n"
+            "```\n"
+            "where `-q:v` generates high-quality JPEG frames and `-start_number 0` asks "
+            "ffmpeg to start the JPEG file from 00000.jpg.")
+
+    frame_names = [p for p in os.listdir(jpg_folder) if os.path.splitext(p)[-1] in [".jpg", ".jpeg", ".JPG", ".JPEG"]]
+    frame_names.sort(key=lambda p: int(os.path.splitext(p)[0]))
+    num_frames = len(frame_names)
+    if num_frames == 0:
+        raise RuntimeError(f"no images found in {jpg_folder}")
+    img_paths = [os.path.join(jpg_folder, frame_name) for frame_name in frame_names]
+    img_mean = torch.tensor(img_mean, dtype=torch.float32)[:, None, None]
+    img_std = torch.tensor(img_std, dtype=torch.float32)[:, None, None]
+
+    if async_loading_frames:
+        lazy_images = AsyncVideoFrameLoader(
+            img_paths,
+            image_size,
+            offload_video_to_cpu,
+            img_mean,
+            img_std,
+            compute_device,
+        )
+        return lazy_images, lazy_images.video_height, lazy_images.video_width
+
+    images = torch.zeros(num_frames, 3, image_size, image_size, dtype=torch.float32)
+    for n, img_path in enumerate(tqdm(img_paths, desc="frame loading (JPEG)")):
+        images[n], video_height, video_width = _load_img_as_tensor(img_path, image_size)
+    if not offload_video_to_cpu:
+        images = images.to(compute_device)
+        img_mean = img_mean.to(compute_device)
+        img_std = img_std.to(compute_device)
+    # normalize by mean and std
+    images -= img_mean
+    images /= img_std
+    return images, video_height, video_width
+
+
+def load_video_frames_from_video_file(
+        video_path,
+        image_size,
+        offload_video_to_cpu,
+        img_mean=(0.485, 0.456, 0.406),
+        img_std=(0.229, 0.224, 0.225),
+        compute_device=torch.device("cuda"),
+):
+    """Load the video frames from a video file."""
+    import decord
+
+    img_mean = torch.tensor(img_mean, dtype=torch.float32)[:, None, None]
+    img_std = torch.tensor(img_std, dtype=torch.float32)[:, None, None]
+    # Get the original video height and width
+    decord.bridge.set_bridge("torch")
+    video_height, video_width, _ = decord.VideoReader(video_path).next().shape
+    # Iterate over all frames in the video
+    images = []
+    for frame in decord.VideoReader(video_path, width=image_size, height=image_size):
+        images.append(frame.permute(2, 0, 1))
+
+    images = torch.stack(images, dim=0).float() / 255.0
+    if not offload_video_to_cpu:
+        images = images.to(compute_device)
+        img_mean = img_mean.to(compute_device)
+        img_std = img_std.to(compute_device)
+    # normalize by mean and std
+    images -= img_mean
+    images /= img_std
+    return images, video_height, video_width
+
+
+def fill_holes_in_mask_scores(mask, max_area):
+    """
+    A post processor to fill small holes in mask scores with area under `max_area`.
+    """
+    # Holes are those connected components in background with area <= self.max_area
+    # (background regions are those with mask scores <= 0)
+    assert max_area > 0, "max_area must be positive"
+
+    input_mask = mask
+    try:
+        labels, areas = get_connected_components(mask <= 0)
+        is_hole = (labels > 0) & (areas <= max_area)
+        # We fill holes with a small positive mask score (0.1) to change them to foreground.
+        mask = torch.where(is_hole, 0.1, mask)
+    except Exception as e:
+        # Skip the post-processing step on removing small holes if the CUDA kernel fails
+        warnings.warn(
+            f"{e}\n\nSkipping the post-processing step due to the error above. You can "
+            "still use SAM 2 and it's OK to ignore the error above, although some post-processing "
+            "functionality may be limited (which doesn't affect the results in most cases; see "
+            "https://github.com/facebookresearch/sam2/blob/main/INSTALL.md).",
+            category=UserWarning,
+            stacklevel=2,
+        )
+        mask = input_mask
+
+    return mask
+
+
+def concat_points(old_point_inputs, new_points, new_labels):
+    """Add new points and labels to previous point inputs (add at the end)."""
+    if old_point_inputs is None:
+        points, labels = new_points, new_labels
+    else:
+        points = torch.cat([old_point_inputs["point_coords"], new_points], dim=1)
+        labels = torch.cat([old_point_inputs["point_labels"], new_labels], dim=1)
+
+    return {"point_coords": points, "point_labels": labels}

sam2/utils/transforms.py

@@ -0,0 +1,108 @@
+# Copyright (c) Meta Platforms, Inc. and affiliates.
+# All rights reserved.
+
+# This source code is licensed under the license found in the
+# LICENSE file in the root directory of this source tree.
+
+import warnings
+
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+from torchvision.transforms import Normalize, Resize, ToTensor
+
+
+class SAM2Transforms(nn.Module):
+
+    def __init__(self, resolution, mask_threshold, max_hole_area=0.0, max_sprinkle_area=0.0):
+        """
+        Transforms for SAM2.
+        """
+        super().__init__()
+        self.resolution = resolution
+        self.mask_threshold = mask_threshold
+        self.max_hole_area = max_hole_area
+        self.max_sprinkle_area = max_sprinkle_area
+        self.mean = [0.485, 0.456, 0.406]
+        self.std = [0.229, 0.224, 0.225]
+        self.to_tensor = ToTensor()
+        self.transforms = torch.jit.script(
+            nn.Sequential(
+                Resize((self.resolution, self.resolution)),
+                Normalize(self.mean, self.std),
+            ))
+
+    def __call__(self, x):
+        x = self.to_tensor(x)
+        return self.transforms(x)
+
+    def forward_batch(self, img_list):
+        img_batch = [self.transforms(self.to_tensor(img)) for img in img_list]
+        img_batch = torch.stack(img_batch, dim=0)
+        return img_batch
+
+    def transform_coords(self, coords: torch.Tensor, normalize=False, orig_hw=None) -> torch.Tensor:
+        """
+        Expects a torch tensor with length 2 in the last dimension. The coordinates can be in absolute image or normalized coordinates,
+        If the coords are in absolute image coordinates, normalize should be set to True and original image size is required.
+
+        Returns
+            Un-normalized coordinates in the range of [0, 1] which is expected by the SAM2 model.
+        """
+        if normalize:
+            assert orig_hw is not None
+            h, w = orig_hw
+            coords = coords.clone()
+            coords[..., 0] = coords[..., 0] / w
+            coords[..., 1] = coords[..., 1] / h
+
+        coords = coords * self.resolution  # unnormalize coords
+        return coords
+
+    def transform_boxes(self, boxes: torch.Tensor, normalize=False, orig_hw=None) -> torch.Tensor:
+        """
+        Expects a tensor of shape Bx4. The coordinates can be in absolute image or normalized coordinates,
+        if the coords are in absolute image coordinates, normalize should be set to True and original image size is required.
+        """
+        boxes = self.transform_coords(boxes.reshape(-1, 2, 2), normalize, orig_hw)
+        return boxes
+
+    def postprocess_masks(self, masks: torch.Tensor, orig_hw) -> torch.Tensor:
+        """
+        Perform PostProcessing on output masks.
+        """
+        from sam2.utils.misc import get_connected_components
+
+        masks = masks.float()
+        input_masks = masks
+        mask_flat = masks.flatten(0, 1).unsqueeze(1)  # flatten as 1-channel image
+        try:
+            if self.max_hole_area > 0:
+                # Holes are those connected components in background with area <= self.fill_hole_area
+                # (background regions are those with mask scores <= self.mask_threshold)
+                labels, areas = get_connected_components(mask_flat <= self.mask_threshold)
+                is_hole = (labels > 0) & (areas <= self.max_hole_area)
+                is_hole = is_hole.reshape_as(masks)
+                # We fill holes with a small positive mask score (10.0) to change them to foreground.
+                masks = torch.where(is_hole, self.mask_threshold + 10.0, masks)
+
+            if self.max_sprinkle_area > 0:
+                labels, areas = get_connected_components(mask_flat > self.mask_threshold)
+                is_hole = (labels > 0) & (areas <= self.max_sprinkle_area)
+                is_hole = is_hole.reshape_as(masks)
+                # We fill holes with negative mask score (-10.0) to change them to background.
+                masks = torch.where(is_hole, self.mask_threshold - 10.0, masks)
+        except Exception as e:
+            # Skip the post-processing step if the CUDA kernel fails
+            warnings.warn(
+                f"{e}\n\nSkipping the post-processing step due to the error above. You can "
+                "still use SAM 2 and it's OK to ignore the error above, although some post-processing "
+                "functionality may be limited (which doesn't affect the results in most cases; see "
+                "https://github.com/facebookresearch/sam2/blob/main/INSTALL.md).",
+                category=UserWarning,
+                stacklevel=2,
+            )
+            masks = input_masks
+
+        masks = F.interpolate(masks.float(), orig_hw, mode="bilinear", align_corners=False).to(masks.dtype)
+        return masks

setup.cfg

@@ -0,0 +1,16 @@
+[yapf]
+column_limit = 120
+based_on_style = pep8
+blank_line_before_nested_class_or_def = true
+split_before_expression_after_opening_paren = true
+
+[isort]
+line_length = 120
+multi_line_output = 0
+known_third_party = cv2,decord,deepspeed,gradio,hydra,imageio,matplotlib,nncore,numpy,omegaconf,pandas,peft,PIL,pycocotools,pysrt,requests,safetensors,spaces,tabulate,termplotlib,tqdm,tensordict,torch,torchvision,transformers
+no_lines_before = STDLIB,LOCALFOLDER
+default_section = FIRSTPARTY
+
+[flake8]
+max-line-length = 500
+extend-ignore = E741

unipixel/constants.py

@@ -0,0 +1,7 @@
+# Copyright (c) 2025 Ye Liu. Licensed under the BSD-3-Clause License.
+
+IGNORE_INDEX = -100
+
+REF_TOKEN = '<|ref|>'
+SEG_TOKEN = '<|seg|>'
+MEM_TOKEN = '<|mem|>'

unipixel/conversation.py

@@ -0,0 +1,49 @@
+# Copyright (c) 2025 Ye Liu. Licensed under the BSD-3-Clause License.
+
+from dataclasses import dataclass
+from typing import List
+
+
+@dataclass
+class Conversation:
+    style: str
+    system: str
+    roles: List[str]
+    seps: List[str]
+    messages: List[str]
+
+    def append_message(self, role, msg):
+        self.messages.append([role, msg])
+
+    def clear(self):
+        self.messages = []
+
+    def get_prompt(self):
+        assert self.style in ('chatml', )
+
+        prompt = self.system + self.seps[0] if self.system is not None else ''
+
+        for i, (role, msg) in enumerate(self.messages):
+            prompt += role
+            sep = self.seps[i % 2]
+            if msg is not None:
+                prompt += msg
+                if not prompt.endswith(sep):
+                    prompt += sep
+
+        prompt = prompt.lstrip('\n')
+        return prompt
+
+
+def get_conv(conv_type):
+    if conv_type == 'chatml':
+        conv = Conversation(
+            style='chatml',
+            system='<|im_start|>system\nYou are a helpful assistant.',
+            roles=('\n<|im_start|>user\n', '\n<|im_start|>assistant\n'),
+            seps=('<|im_end|>', '<|im_end|>'),
+            messages=[])
+    else:
+        raise ValueError(f'unknown conversation type: {conv_type}')
+
+    return conv

unipixel/dataset/utils.py

@@ -0,0 +1,531 @@
+# Copyright (c) 2025 Ye Liu. Licensed under the BSD-3-Clause License.
+
+import base64
+import copy
+import math
+import os
+import warnings
+from io import BytesIO
+from typing import Optional
+
+import cv2
+import decord
+import nncore
+import numpy as np
+import requests
+import torch
+import torchvision.transforms.functional as T
+from PIL import Image
+from pycocotools.mask import decode, frPyObjects, merge
+from torchvision import transforms
+from torchvision.transforms import InterpolationMode
+
+from unipixel.constants import IGNORE_INDEX
+from unipixel.conversation import get_conv
+
+IMAGE_FACTOR = 28
+MIN_PIXELS = 4 * 28 * 28
+MAX_PIXELS = 16384 * 28 * 28
+MAX_RATIO = 200
+
+VIDEO_MIN_PIXELS = 128 * 28 * 28
+VIDEO_MAX_PIXELS = 768 * 28 * 28
+FRAME_FACTOR = 2
+FPS = 2.0
+FPS_MIN_FRAMES = 4
+FPS_MAX_FRAMES = 768
+
+# Set the maximum number of video token inputs.
+# Here, 128K represents the maximum number of input tokens for the VLLM model.
+# Remember to adjust it according to your own configuration.
+VIDEO_TOTAL_PIXELS = int(float(os.environ.get('VIDEO_MAX_PIXELS', 128000 * 28 * 28 * 0.9)))
+
+
+def round_by_factor(number: int, factor: int) -> int:
+    """Returns the closest integer to 'number' that is divisible by 'factor'."""
+    return round(number / factor) * factor
+
+
+def ceil_by_factor(number: int, factor: int) -> int:
+    """Returns the smallest integer greater than or equal to 'number' that is divisible by 'factor'."""
+    return math.ceil(number / factor) * factor
+
+
+def floor_by_factor(number: int, factor: int) -> int:
+    """Returns the largest integer less than or equal to 'number' that is divisible by 'factor'."""
+    return math.floor(number / factor) * factor
+
+
+def smart_resize(height: int,
+                 width: int,
+                 factor: int = IMAGE_FACTOR,
+                 min_pixels: int = MIN_PIXELS,
+                 max_pixels: int = MAX_PIXELS) -> tuple[int, int]:
+    """
+    Rescales the image so that the following conditions are met:
+
+    1. Both dimensions (height and width) are divisible by 'factor'.
+
+    2. The total number of pixels is within the range ['min_pixels', 'max_pixels'].
+
+    3. The aspect ratio of the image is maintained as closely as possible.
+    """
+    if max(height, width) / min(height, width) > MAX_RATIO:
+        raise ValueError(
+            f"absolute aspect ratio must be smaller than {MAX_RATIO}, got {max(height, width) / min(height, width)}")
+    h_bar = max(factor, round_by_factor(height, factor))
+    w_bar = max(factor, round_by_factor(width, factor))
+    # change order here to ensure not exceeding max_pixels
+    if h_bar * w_bar < min_pixels:
+        beta = math.sqrt(min_pixels / (height * width))
+        h_bar = ceil_by_factor(height * beta, factor)
+        w_bar = ceil_by_factor(width * beta, factor)
+    if h_bar * w_bar > max_pixels:
+        beta = math.sqrt((height * width) / max_pixels)
+        h_bar = floor_by_factor(height / beta, factor)
+        w_bar = floor_by_factor(width / beta, factor)
+    return h_bar, w_bar
+
+
+def to_rgb(pil_image: Image.Image) -> Image.Image:
+    if pil_image.mode == 'RGBA':
+        white_background = Image.new("RGB", pil_image.size, (255, 255, 255))
+        white_background.paste(pil_image, mask=pil_image.split()[3])  # Use alpha channel as mask
+        return white_background
+    else:
+        return pil_image.convert("RGB")
+
+
+def fetch_image(ele: dict[str, str | Image.Image], size_factor: int = IMAGE_FACTOR) -> Image.Image:
+    if "image" in ele:
+        image = ele["image"]
+    else:
+        image = ele["image_url"]
+    image_obj = None
+    if isinstance(image, Image.Image):
+        image_obj = image
+    elif image.startswith("http://") or image.startswith("https://"):
+        # fix memory leak issue while using BytesIO
+        with requests.get(image, stream=True) as response:
+            response.raise_for_status()
+            with BytesIO(response.content) as bio:
+                image_obj = copy.deepcopy(Image.open(bio))
+    elif image.startswith("file://"):
+        image_obj = Image.open(image[7:])
+    elif image.startswith("data:image"):
+        if "base64," in image:
+            _, base64_data = image.split("base64,", 1)
+            data = base64.b64decode(base64_data)
+            # fix memory leak issue while using BytesIO
+            with BytesIO(data) as bio:
+                image_obj = copy.deepcopy(Image.open(bio))
+    else:
+        image_obj = Image.open(image)
+    if image_obj is None:
+        raise ValueError(f"Unrecognized image input, support local path, http url, base64 and PIL.Image, got {image}")
+    image = to_rgb(image_obj)
+
+    if "resized_height" in ele and "resized_width" in ele:
+        resized_height, resized_width = smart_resize(
+            ele["resized_height"],
+            ele["resized_width"],
+            factor=size_factor,
+        )
+    else:
+        width, height = image.size
+        min_pixels = ele.get("min_pixels", MIN_PIXELS)
+        max_pixels = ele.get("max_pixels", MAX_PIXELS)
+        resized_height, resized_width = smart_resize(
+            height,
+            width,
+            factor=size_factor,
+            min_pixels=min_pixels,
+            max_pixels=max_pixels,
+        )
+    image = image.resize((resized_width, resized_height))
+
+    return image
+
+
+def smart_nframes(
+    ele: dict,
+    total_frames: int,
+    video_fps: int | float,
+) -> int:
+    """calculate the number of frames for video used for model inputs.
+
+    Args:
+        ele (dict): a dict contains the configuration of video.
+            support either `fps` or `nframes`:
+                - nframes: the number of frames to extract for model inputs.
+                - fps: the fps to extract frames for model inputs.
+                    - min_frames: the minimum number of frames of the video, only used when fps is provided.
+                    - max_frames: the maximum number of frames of the video, only used when fps is provided.
+        total_frames (int): the original total number of frames of the video.
+        video_fps (int | float): the original fps of the video.
+
+    Raises:
+        ValueError: nframes should in interval [FRAME_FACTOR, total_frames].
+
+    Returns:
+        int: the number of frames for video used for model inputs.
+    """
+    assert not ("fps" in ele and "nframes" in ele), "Only accept either `fps` or `nframes`"
+    if "nframes" in ele:
+        nframes = round_by_factor(ele["nframes"], FRAME_FACTOR)
+    else:
+        fps = ele.get("fps", FPS)
+        min_frames = ceil_by_factor(ele.get("min_frames", FPS_MIN_FRAMES), FRAME_FACTOR)
+        max_frames = floor_by_factor(ele.get("max_frames", min(FPS_MAX_FRAMES, total_frames)), FRAME_FACTOR)
+        nframes = total_frames / video_fps * fps
+        nframes = min(min(max(nframes, min_frames), max_frames), total_frames)
+        nframes = floor_by_factor(nframes, FRAME_FACTOR)
+    if not (FRAME_FACTOR <= nframes and nframes <= total_frames):
+        raise ValueError(f"nframes should in interval [{FRAME_FACTOR}, {total_frames}], but got {nframes}.")
+    return nframes
+
+
+def calculate_video_frame_range(
+    ele: dict,
+    total_frames: int,
+    video_fps: float,
+) -> tuple[int, int, int]:
+    """
+    Calculate the start and end frame indices based on the given time range.
+
+    Args:
+        ele (dict): A dictionary containing optional 'video_start' and 'video_end' keys (in seconds).
+        total_frames (int): Total number of frames in the video.
+        video_fps (float): Frames per second of the video.
+
+    Returns:
+        tuple: A tuple containing (start_frame, end_frame, frame_count).
+
+    Raises:
+        ValueError: If input parameters are invalid or the time range is inconsistent.
+    """
+    # Validate essential parameters
+    if video_fps <= 0:
+        raise ValueError("video_fps must be a positive number")
+    if total_frames <= 0:
+        raise ValueError("total_frames must be a positive integer")
+
+    # Get start and end time in seconds
+    video_start = ele.get("video_start", None)
+    video_end = ele.get("video_end", None)
+    if video_start is None and video_end is None:
+        return 0, total_frames - 1, total_frames
+
+    max_duration = total_frames / video_fps
+    # Process start frame
+    if video_start is not None:
+        video_start_clamped = max(0.0, min(video_start, max_duration))
+        start_frame = math.ceil(video_start_clamped * video_fps)
+    else:
+        start_frame = 0
+    # Process end frame
+    if video_end is not None:
+        video_end_clamped = max(0.0, min(video_end, max_duration))
+        end_frame = math.floor(video_end_clamped * video_fps)
+        end_frame = min(end_frame, total_frames - 1)
+    else:
+        end_frame = total_frames - 1
+
+    # Validate frame order
+    if start_frame >= end_frame:
+        raise ValueError(
+            f"Invalid time range: Start frame {start_frame} (at {video_start_clamped if video_start is not None else 0}s) "
+            f"exceeds end frame {end_frame} (at {video_end_clamped if video_end is not None else max_duration}s). "
+            f"Video duration: {max_duration:.2f}s ({total_frames} frames @ {video_fps}fps)")
+
+    return start_frame, end_frame, end_frame - start_frame + 1
+
+
+def _read_video_decord(ele: dict, ) -> (torch.Tensor, float):
+    """read video using decord.VideoReader
+
+    Args:
+        ele (dict): a dict contains the configuration of video.
+        support keys:
+            - video: the path of video. support "file://", "http://", "https://" and local path.
+            - video_start: the start time of video.
+            - video_end: the end time of video.
+    Returns:
+        torch.Tensor: the video tensor with shape (T, C, H, W).
+    """
+    decord.bridge.set_bridge("torch")
+    video_path = ele["video"]
+    vr = decord.VideoReader(video_path, num_threads=ele.get('num_threads', 0))
+    total_frames, video_fps = len(vr), vr.get_avg_fps()
+    start_frame, end_frame, total_frames = calculate_video_frame_range(
+        ele,
+        total_frames,
+        video_fps,
+    )
+    nframes = smart_nframes(ele, total_frames=total_frames, video_fps=video_fps)
+    idx = torch.linspace(start_frame, end_frame, nframes).round().long().tolist()
+    video = vr.get_batch(idx).permute(0, 3, 1, 2)  # Convert to TCHW format
+    sample_fps = nframes / max(total_frames, 1e-6) * video_fps
+    return video, sample_fps
+
+
+def fetch_video(ele: dict,
+                image_factor: int = IMAGE_FACTOR,
+                return_video_sample_fps: bool = False,
+                sanity_check=False) -> torch.Tensor | list[Image.Image]:
+    if isinstance(ele["video"], str):
+        video, sample_fps = _read_video_decord(ele)
+        nframes, _, height, width = video.shape
+        min_pixels = ele.get("min_pixels", VIDEO_MIN_PIXELS)
+        total_pixels = ele.get("total_pixels", VIDEO_TOTAL_PIXELS)
+        max_pixels = max(min(VIDEO_MAX_PIXELS, total_pixels / nframes * FRAME_FACTOR), int(min_pixels * 1.05))
+        max_pixels_supposed = ele.get("max_pixels", max_pixels)
+        max_pixels = min(max_pixels_supposed, max_pixels)
+        if "resized_height" in ele and "resized_width" in ele:
+            resized_height, resized_width = smart_resize(
+                ele["resized_height"],
+                ele["resized_width"],
+                factor=image_factor,
+            )
+        else:
+            resized_height, resized_width = smart_resize(
+                height,
+                width,
+                factor=image_factor,
+                min_pixels=min_pixels,
+                max_pixels=max_pixels,
+            )
+        video = transforms.functional.resize(
+            video,
+            [resized_height, resized_width],
+            interpolation=InterpolationMode.BICUBIC,
+            antialias=True,
+        ).float()
+
+        if sanity_check and (video == 0).all():
+            raise ValueError("video '{}' contains all zeros".format(ele["video"]))
+
+        if return_video_sample_fps:
+            return video, sample_fps
+        return video
+    else:
+        assert isinstance(ele["video"], (list, tuple))
+        process_info = ele.copy()
+        process_info.pop("type", None)
+        process_info.pop("video", None)
+        images = [
+            fetch_image({
+                "image": video_element,
+                **process_info
+            }, size_factor=image_factor) for video_element in ele["video"]
+        ]
+        nframes = ceil_by_factor(len(images), FRAME_FACTOR)
+        if len(images) < nframes:
+            images.extend([images[-1]] * (nframes - len(images)))
+        if return_video_sample_fps:
+            return images, process_info.pop("fps", 2.0)
+        return images
+
+
+def extract_vision_info(conversations: list[dict] | list[list[dict]]) -> list[dict]:
+    vision_infos = []
+    if isinstance(conversations[0], dict):
+        conversations = [conversations]
+    for conversation in conversations:
+        for message in conversation:
+            if isinstance(message["content"], list):
+                for ele in message["content"]:
+                    if ("image" in ele or "image_url" in ele or "video" in ele
+                            or ele.get("type", "") in ("image", "image_url", "video")):
+                        vision_infos.append(ele)
+    return vision_infos
+
+
+def process_vision_info(
+    conversations: list[dict] | list[list[dict]],
+    return_video_kwargs: bool = False,
+    sanity_check=False
+) -> tuple[list[Image.Image] | None, list[torch.Tensor | list[Image.Image]] | None, Optional[dict]]:
+
+    vision_infos = extract_vision_info(conversations)
+    # Read images or videos
+    image_inputs = []
+    video_inputs = []
+    video_sample_fps_list = []
+    for vision_info in vision_infos:
+        if "image" in vision_info or "image_url" in vision_info:
+            image_inputs.append(fetch_image(vision_info))
+        elif "video" in vision_info:
+            video_input, video_sample_fps = fetch_video(
+                vision_info, return_video_sample_fps=True, sanity_check=sanity_check)
+            video_sample_fps_list.append(video_sample_fps)
+            video_inputs.append(video_input)
+        else:
+            raise ValueError("image, image_url or video should in content.")
+    if len(image_inputs) == 0:
+        image_inputs = None
+    if len(video_inputs) == 0:
+        video_inputs = None
+    if return_video_kwargs:
+        return image_inputs, video_inputs, {'fps': video_sample_fps_list}
+    return image_inputs, video_inputs
+
+
+def resize(mask, size):
+    return T.resize(mask.unsqueeze(0).unsqueeze(0), size)[0, 0]
+
+
+def process_masks(sample, frame_size, inds):
+    if sample['mask_type'] == 'image':
+        # case 1: list of masks or paths to masks
+        masks = []
+        for obj_oids in sample['oids']:
+            obj_masks = []
+            for i in inds:
+                label = sample['masks'][i]
+                if isinstance(label, str):
+                    label = np.array(Image.open(label))
+                elif label is None:
+                    label = np.full(frame_size, -1)
+                obj_masks.append(torch.from_numpy(sum([label == oid for oid in obj_oids])).float())
+            masks.append(obj_masks)
+    elif sample['mask_type'] == 'image_sep':
+        # case 2: list of masks or paths to masks (one object per image)
+        masks = []
+        for raw_obj_masks in sample['masks']:
+            obj_masks = []
+            for i in inds:
+                label = raw_obj_masks[i]
+                if isinstance(label, str):
+                    label = np.array(Image.open(label))
+                elif label is None:
+                    label = np.full(frame_size, -1)
+                obj_masks.append(torch.from_numpy(label == 255).float())
+            masks.append(obj_masks)
+    elif sample['mask_type'] == 'rle':
+        # case 3: list of lists of multi-region RLE masks
+        raw_masks = nncore.load(sample['masks']) if isinstance(sample['masks'], str) else sample['masks']
+        masks = []
+        for raw_obj_masks in raw_masks:
+            obj_masks = []
+            for i in inds:
+                mask = torch.zeros(frame_size)
+                for rle in raw_obj_masks[i]:
+                    if isinstance(rle, list):
+                        rles = frPyObjects(rle, sample.get('height', frame_size[0]), sample.get('width', frame_size[1]))
+                        mask += resize(torch.from_numpy(decode(merge(rles))).float(), frame_size)
+                    elif isinstance(rle, dict):
+                        if isinstance(rle['counts'], list):
+                            rle = frPyObjects(rle, *rle['size'])
+                        mask += resize(torch.from_numpy(decode(rle)).float(), frame_size)
+                    elif rle is None:
+                        mask += 0
+                    else:
+                        raise TypeError(f'unknown rle mask: {rle}')
+                obj_masks.append((mask > 0).float())
+            masks.append(obj_masks)
+    elif sample['mask_type'] == 'polygon':
+        # case 4: list of lists of polygons
+        masks = []
+        for raw_obj_masks in sample['masks']:
+            obj_masks = []
+            for i in inds:
+                # step 1: sort shapes
+                areas = []
+                for shape in raw_obj_masks[i]:
+                    tmp = np.zeros(frame_size, dtype=np.uint8)
+                    cv2.polylines(tmp, np.array([shape['points']], dtype=np.int32), True, 1, 1)
+                    cv2.fillPoly(tmp, np.array([shape['points']], dtype=np.int32), 1)
+                    areas.append(tmp.sum())
+                shapes = [raw_obj_masks[i][j] for j in list(np.argsort(areas)[::-1].astype(np.int32))]
+                # step 2: draw masks
+                mask = np.zeros(frame_size, dtype=np.uint8)
+                for shape in shapes:
+                    assert shape['label'] in ('target', 'ignore'), shape
+                    label = 1 if shape['label'] == 'target' else -1  # replacing 255 with -1 here
+                    cv2.polylines(mask, np.array([shape['points']], dtype=np.int32), True, label, 1)
+                    cv2.fillPoly(mask, np.array([shape['points']], dtype=np.int32), label)
+                obj_masks.append(torch.from_numpy(mask).float())
+            masks.append(obj_masks)
+    elif sample['mask_type'] == 'vicas':
+        # case 5: special case for vicas dataset
+        masks = []
+        for obj_rle_path in sample['masks']:
+            obj_rles, obj_masks = nncore.load(obj_rle_path), []
+            for i in inds:
+                mask = torch.zeros(frame_size)
+                for rle in obj_rles[i]:
+                    mask += 0 if rle is None else resize(torch.from_numpy(decode(rle)).float(), frame_size)
+                obj_masks.append((mask > 0).float())
+            masks.append(obj_masks)
+    elif sample['mask_type'] == 'sav':
+        # case 6: special case for sav dataset
+        annos = nncore.load(sample['masks'])['masklet']
+        masks = [[]]
+        for i in inds:
+            mask = resize(torch.from_numpy(decode(annos[i][int(sample['qid'])])).float(), frame_size)
+            masks[0].append(mask)
+    else:
+        raise TypeError(f"unknown mask type: {sample['mask_type']}")
+
+    return masks
+
+
+def build_obj_to_frame_idx(label_mask, batch_mode):
+    step_t_obj_to_frame_idx = [[]] if batch_mode else [[] for _ in range(label_mask.size(0))]
+
+    # t: frame_idx v: video_idx
+    for t in range(len(step_t_obj_to_frame_idx)):
+        if batch_mode:
+            for v in range(label_mask.size(0)):
+                for _ in range(label_mask.size(1)):
+                    step_t_obj_to_frame_idx[t].append(torch.IntTensor([t, v]))
+        else:
+            for _ in range(label_mask.size(1)):
+                step_t_obj_to_frame_idx[t].append(torch.IntTensor([t, 0]))
+
+    label_obj_to_frame_idx = torch.stack([torch.stack(o) for o in step_t_obj_to_frame_idx])
+    return label_obj_to_frame_idx
+
+
+def preprocess_chatml(input_ids, text, tokenizer):
+    conv = get_conv('chatml')
+
+    rounds = [m + conv.seps[0] for m in text.split(conv.seps[0])]
+    assert (len(rounds) % 2 == 0) == (conv.system is not None)
+    assert rounds[-1] == conv.seps[0]
+    rounds = rounds[:-1]
+
+    if conv.system is None:
+        rounds = [''.join(rounds[i:i + 2]) for i in range(0, len(rounds), 2)]
+    else:
+        rounds = [''.join(rounds[:3])] + [''.join(rounds[i:i + 2]) for i in range(3, len(rounds), 2)]
+
+    labels = input_ids.clone()
+
+    sep = conv.seps[0] + conv.roles[1]
+    cur_len = 0
+
+    for i, rou in enumerate(rounds):
+        if len(rou) == 0:
+            break
+
+        ins = sep.join(rou.split(sep)[:-1]) + sep
+
+        rou_len = tokenizer(rou, return_length=True).length[0]
+        ins_len = tokenizer(ins, return_length=True).length[0]
+
+        labels[cur_len:cur_len + ins_len] = IGNORE_INDEX
+        cur_len += rou_len
+
+    if labels.size(0) != cur_len:
+        warnings.warn(f'Tokenization mismatch: {labels.size(0)} and {cur_len}')
+
+    return labels
+
+
+def preprocess(input_ids, text, tokenizer, conv_type):
+    if conv_type == 'chatml':
+        return preprocess_chatml(input_ids, text, tokenizer)
+    else:
+        raise ValueError(f'unknown conversation type: {conv_type}')

unipixel/model/__init__.py

@@ -0,0 +1,5 @@
+# Copyright (c) 2025 Ye Liu. Licensed under the BSD-3-Clause License.
+
+from .qwen2_5_vl import PatchedQwen2_5_VLProcessor, PixelQwen2_5_VLConfig, PixelQwen2_5_VLForConditionalGeneration
+
+MODELS = {'qwen2_5_vl': (PixelQwen2_5_VLConfig, PixelQwen2_5_VLForConditionalGeneration, PatchedQwen2_5_VLProcessor)}

unipixel/model/builder.py

@@ -0,0 +1,109 @@
+# Copyright (c) 2025 Ye Liu. Licensed under the BSD-3-Clause License.
+
+import nncore
+import torch
+import torch.nn as nn
+from peft import PeftModel
+from safetensors.torch import load_model
+from transformers import AutoConfig, AutoModel, AutoProcessor, Qwen2_5_VLForConditionalGeneration
+
+from unipixel.utils.env import get_auto_device
+
+
+def build_model(model_path,
+                config=None,
+                image_size=None,
+                is_trainable=False,
+                merge_adapter=False,
+                attn_implementation='flash_attention_2',
+                device='auto',
+                dtype='bfloat16'):
+    # set do_resize to false to avoid duplicated resizing
+    # https://github.com/huggingface/transformers/blob/main/src/transformers/models/qwen2_vl/image_processing_qwen2_vl.py
+    processor = AutoProcessor.from_pretrained(model_path, use_fast=True, do_resize=False)
+
+    config = config or AutoConfig.from_pretrained(model_path)
+    config.sam2_inference_mode = not is_trainable
+
+    # override sam2 image size
+    if image_size is not None:
+        config.sam2_image_size = image_size
+
+    adapter_path = nncore.join(model_path, 'adapter_model.safetensors')
+    partial_path = nncore.join(model_path, 'pytorch_model.safetensors')
+
+    if nncore.is_file(adapter_path) or nncore.is_file(partial_path):
+        print(f'Loading base model from {config.base_model_path}...')
+        model = AutoModel.from_pretrained(
+            config.base_model_path,
+            config=config,
+            low_cpu_mem_usage=True,
+            ignore_mismatched_sizes=True,
+            attn_implementation=attn_implementation,
+            torch_dtype=dtype,
+            device_map='auto' if device == 'all' else None)
+
+        meta_state_dict = {
+            n: torch.empty_like(p, device='cpu')
+            for n, p in model.named_parameters() if p.device == torch.device('meta')
+        }
+        model.load_state_dict(meta_state_dict, strict=False, assign=True)
+
+        # sam2 weights might be replaced later
+        if model.config.sam2_checkpoint:
+            model.load_sam2_weights()
+
+        embed_tokens = model.get_input_embeddings()
+        size = (embed_tokens.num_embeddings, embed_tokens.embedding_dim)
+        if embed_tokens.weight.size() != size:
+            print(f'Resizing embed_tokens from {embed_tokens.weight.size()} to {size}...')
+            model.model.language_model.embed_tokens.weight = nn.Parameter(embed_tokens.weight.new_empty(size))
+
+        size = (model.lm_head.out_features, model.lm_head.in_features)
+        if model.lm_head.weight.size() != size:
+            print(f'Resizing lm_head from {model.lm_head.weight.size()} to {size}...')
+            model.lm_head.weight = nn.Parameter(model.lm_head.weight.new_empty(size))
+
+        if nncore.is_file(adapter_path):
+            print(f'Loading adapter from {model_path}...')
+            # transformers integration does not support merge_and_unload, use peft instead
+            model = PeftModel.from_pretrained(
+                model,
+                model_path,
+                is_trainable=is_trainable,
+                low_cpu_mem_usage=True,
+                # load adapters to the same device as embed_tokens
+                torch_device=str(embed_tokens.weight.device))
+
+        if nncore.is_file(partial_path):
+            print(f'Loading state dict from {partial_path}...')
+            _, unexpected = load_model(model, partial_path, strict=False, device=str(model.device))
+            assert len(unexpected) == 0, f'unexpected parameters: {unexpected}'
+
+        if (not is_trainable or merge_adapter) and nncore.is_file(adapter_path):
+            print('Merging adapter and unloading...')
+            model = model.merge_and_unload()
+            model._hf_peft_config_loaded = False
+    else:
+        print(f'Loading full model from {model_path}...')
+
+        if config.model_type == 'qwen2_5_vl':
+            model_cls = Qwen2_5_VLForConditionalGeneration
+        else:
+            model_cls = AutoModel
+
+        model = model_cls.from_pretrained(
+            model_path,
+            config=config,
+            low_cpu_mem_usage=True,
+            attn_implementation=attn_implementation,
+            torch_dtype=dtype,
+            device_map='auto' if device == 'all' else None)
+
+        model.requires_grad_(False)
+
+    if not is_trainable and device != 'all':
+        device = get_auto_device() if device == 'auto' else device
+        model = model.to(device).eval()
+
+    return model, processor