Upload 8 files

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* 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
+example.jpg filter=lfs diff=lfs merge=lfs -text

config.json

@@ -0,0 +1,142 @@
+{
+  "_name_or_path": "facebook/detr-resnet-50",
+  "activation_dropout": 0.0,
+  "activation_function": "relu",
+  "architectures": [
+    "DetrForObjectDetection"
+  ],
+  "attention_dropout": 0.0,
+  "auxiliary_loss": false,
+  "backbone": null,
+  "backbone_config": {
+    "_name_or_path": "",
+    "add_cross_attention": false,
+    "architectures": null,
+    "bad_words_ids": null,
+    "begin_suppress_tokens": null,
+    "bos_token_id": null,
+    "chunk_size_feed_forward": 0,
+    "cross_attention_hidden_size": null,
+    "decoder_start_token_id": null,
+    "depths": [
+      3,
+      4,
+      6,
+      3
+    ],
+    "diversity_penalty": 0.0,
+    "do_sample": false,
+    "downsample_in_first_stage": false,
+    "early_stopping": false,
+    "embedding_size": 64,
+    "encoder_no_repeat_ngram_size": 0,
+    "eos_token_id": null,
+    "exponential_decay_length_penalty": null,
+    "finetuning_task": null,
+    "forced_bos_token_id": null,
+    "forced_eos_token_id": null,
+    "hidden_act": "relu",
+    "hidden_sizes": [
+      256,
+      512,
+      1024,
+      2048
+    ],
+    "id2label": {
+      "0": "LABEL_0",
+      "1": "LABEL_1"
+    },
+    "is_decoder": false,
+    "is_encoder_decoder": false,
+    "label2id": {
+      "LABEL_0": 0,
+      "LABEL_1": 1
+    },
+    "layer_type": "bottleneck",
+    "length_penalty": 1.0,
+    "max_length": 20,
+    "min_length": 0,
+    "model_type": "resnet",
+    "no_repeat_ngram_size": 0,
+    "num_beam_groups": 1,
+    "num_beams": 1,
+    "num_channels": 3,
+    "num_return_sequences": 1,
+    "out_features": [
+      "stage4"
+    ],
+    "out_indices": [
+      4
+    ],
+    "output_attentions": false,
+    "output_hidden_states": false,
+    "output_scores": false,
+    "pad_token_id": null,
+    "prefix": null,
+    "problem_type": null,
+    "pruned_heads": {},
+    "remove_invalid_values": false,
+    "repetition_penalty": 1.0,
+    "return_dict": true,
+    "return_dict_in_generate": false,
+    "sep_token_id": null,
+    "stage_names": [
+      "stem",
+      "stage1",
+      "stage2",
+      "stage3",
+      "stage4"
+    ],
+    "suppress_tokens": null,
+    "task_specific_params": null,
+    "temperature": 1.0,
+    "tf_legacy_loss": false,
+    "tie_encoder_decoder": false,
+    "tie_word_embeddings": true,
+    "tokenizer_class": null,
+    "top_k": 50,
+    "top_p": 1.0,
+    "torch_dtype": null,
+    "torchscript": false,
+    "typical_p": 1.0,
+    "use_bfloat16": false
+  },
+  "bbox_cost": 5,
+  "bbox_loss_coefficient": 5,
+  "class_cost": 1,
+  "d_model": 256,
+  "decoder_attention_heads": 8,
+  "decoder_ffn_dim": 2048,
+  "decoder_layerdrop": 0.0,
+  "decoder_layers": 6,
+  "dice_loss_coefficient": 1,
+  "dilation": null,
+  "dropout": 0.1,
+  "encoder_attention_heads": 8,
+  "encoder_ffn_dim": 2048,
+  "encoder_layerdrop": 0.0,
+  "encoder_layers": 6,
+  "eos_coefficient": 0.1,
+  "giou_cost": 2,
+  "giou_loss_coefficient": 2,
+  "id2label": {
+    "0": "LABEL_0"
+  },
+  "init_std": 0.02,
+  "init_xavier_std": 1.0,
+  "is_encoder_decoder": true,
+  "label2id": {
+    "LABEL_0": 0
+  },
+  "mask_loss_coefficient": 1,
+  "model_type": "detr",
+  "num_channels": 3,
+  "num_hidden_layers": 6,
+  "num_queries": 100,
+  "position_embedding_type": "sine",
+  "scale_embedding": false,
+  "torch_dtype": "float32",
+  "transformers_version": "4.34.0",
+  "use_pretrained_backbone": null,
+  "use_timm_backbone": false
+}

element.yaml

@@ -0,0 +1,6 @@
+version: 0.1.0
+element_type: Detect
+main: main.py
+source: https://huggingface.co/journey2001/qr_detect_300ep
+objects:
+  - qr_code

main.py

@@ -0,0 +1,111 @@
+from __future__ import annotations
+
+import argparse
+import json
+import sys
+from io import BytesIO
+from pathlib import Path
+from typing import Any, Dict, List, Tuple
+
+import numpy as np
+from PIL import Image
+import torch
+from transformers import AutoImageProcessor, DetrForObjectDetection
+
+
+ModelBundle = Tuple[DetrForObjectDetection, AutoImageProcessor]
+
+
+def load_image(frame: Any, base_dir: Path) -> Image.Image:
+    if isinstance(frame, (bytes, bytearray, memoryview)):
+        return Image.open(BytesIO(frame)).convert("RGB")
+
+    path = Path(str(frame))
+    if not path.is_absolute():
+        path = (Path.cwd() / path).resolve()
+        if not path.exists():
+            candidate = (base_dir / str(frame)).resolve()
+            if candidate.exists():
+                path = candidate
+    return Image.open(path).convert("RGB")
+
+
+def load_model(*_args: Any, **_kwargs: Any) -> ModelBundle | None:
+    base_dir = Path(__file__).resolve().parent
+    if not (base_dir / "config.json").exists():
+        return None
+
+    processor = AutoImageProcessor.from_pretrained(str(base_dir))
+    model = DetrForObjectDetection.from_pretrained(str(base_dir))
+    model.eval()
+    return model, processor
+
+
+def run_model(model_bundle: ModelBundle, frame: "np.ndarray") -> List[Dict[str, Any]]:
+    image = Image.fromarray(frame)
+    model, processor = model_bundle
+
+    inputs = processor(images=image, return_tensors="pt")
+    with torch.no_grad():
+        outputs = model(**inputs)
+
+    target_sizes = torch.tensor([image.size[::-1]])
+    results = processor.post_process_object_detection(
+        outputs,
+        threshold=0.5,
+        target_sizes=target_sizes,
+    )[0]
+
+    detections: List[Dict[str, Any]] = []
+    names = model.config.id2label or {}
+    name_overrides = {"LABEL_0": "qr_code"}
+
+    for det_idx, (score, label, box) in enumerate(
+        zip(results["scores"], results["labels"], results["boxes"])
+    ):
+        class_id = int(label.item())
+        class_name = names.get(class_id, str(class_id))
+        class_name = name_overrides.get(class_name, class_name)
+        xyxy = [float(x) for x in box.tolist()]
+        detections.append(
+            {
+                "frame_idx": 0,
+                "class": class_name,
+                "bbox": xyxy,
+                "score": float(score.item()),
+                "track_id": f"f0-d{det_idx}",
+            }
+        )
+
+    return detections
+
+
+def build_parser() -> argparse.ArgumentParser:
+    parser = argparse.ArgumentParser(description="Run QR detect element.")
+    parser.add_argument(
+        "--stdin-raw",
+        action="store_true",
+        default=True,
+        help="Read raw image bytes from stdin.",
+    )
+    return parser
+
+
+if __name__ == "__main__":
+    build_parser().parse_args()
+
+    base_dir = Path(__file__).resolve().parent
+    model_bundle = load_model()
+    if model_bundle is None:
+        print("[]")
+        sys.exit(0)
+
+    try:
+        image = load_image(sys.stdin.buffer.read(), base_dir)
+    except Exception:
+        print("[]")
+        sys.exit(0)
+
+    frame = np.array(image)
+    output = run_model(model_bundle, frame)
+    print(json.dumps(output))

preprocessor_config.json

@@ -0,0 +1,25 @@
+{
+  "do_normalize": true,
+  "do_pad": true,
+  "do_rescale": true,
+  "do_resize": true,
+  "feature_extractor_type": "DetrFeatureExtractor",
+  "format": "coco_detection",
+  "image_mean": [
+    0.485,
+    0.456,
+    0.406
+  ],
+  "image_processor_type": "DetrImageProcessor",
+  "image_std": [
+    0.229,
+    0.224,
+    0.225
+  ],
+  "resample": 2,
+  "rescale_factor": 0.00392156862745098,
+  "size": {
+    "longest_edge": 1333,
+    "shortest_edge": 800
+  }
+}

pyproject.toml

@@ -0,0 +1,10 @@
+[project]
+name = "qr-detect-300ep"
+version = "0.1.0"
+requires-python = ">=3.11"
+dependencies = [
+  "numpy>=1.26",
+  "pillow>=10.0",
+  "torch>=2.2",
+  "transformers>=4.38",
+]

pytorch_model.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3b5eff261ebe4f1048560c50df8bfe1584aff25870e0fff4035e8c829b7c8358
+size 166616594