File size: 8,296 Bytes
3d1c0e1
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
import os
import sys
sys.path.append(os.getcwd())

import argparse
import warnings
warnings.filterwarnings("ignore")

import time
from contextlib import nullcontext
from omegaconf import OmegaConf
from pathlib import Path
from tqdm import tqdm

import numpy as np
import torch
import decord
from einops import rearrange
from lightning.pytorch import seed_everything
from torch import autocast
from torchvision import transforms
from torchvision.io import write_video

from vidtok.modules.util import print0
from scripts.inference_evaluate import load_model_from_config


class SingleVideoDataset(torch.utils.data.Dataset):
    def __init__(
        self, 
        video_path, 
        input_height=128, 
        input_width=128, 
        sample_fps=8,
        chunk_size=16,
        is_causal=True,
        read_long_video=False
    ):
        decord.bridge.set_bridge("torch")
        self.video_path = video_path
        self.transform = transforms.Compose(
            [
                transforms.Resize(input_height, antialias=True),
                transforms.CenterCrop((input_height, input_width)),
                transforms.Normalize(mean=(0.5, 0.5, 0.5), std=(0.5, 0.5, 0.5)),
            ]
        )

        self.video_reader = decord.VideoReader(video_path, num_threads=0)
        total_frames = len(self.video_reader)
        fps = self.video_reader.get_avg_fps()  # float

        interval = round(fps / sample_fps)
        frame_ids = list(range(0, total_frames, interval))
        self.frame_ids_batch = []
        if read_long_video:
            video_length = len(frame_ids)
            if is_causal and video_length > chunk_size:
                self.frame_ids_batch.append(frame_ids[:chunk_size * ((video_length - 1) // chunk_size) + 1])
            elif not is_causal and video_length >= chunk_size:
                self.frame_ids_batch.append(frame_ids[:chunk_size * (video_length // chunk_size)])
        else:
            num_frames_per_batch = chunk_size + 1 if is_causal else chunk_size
            for x in range(0, len(frame_ids), num_frames_per_batch):
                if len(frame_ids[x : x + num_frames_per_batch]) == num_frames_per_batch:
                    self.frame_ids_batch.append(frame_ids[x : x + num_frames_per_batch])

    def __len__(self):
        return len(self.frame_ids_batch)

    def __getitem__(self, idx):
        frame_ids = self.frame_ids_batch[idx]
        frames = self.video_reader.get_batch(frame_ids).permute(0, 3, 1, 2).float() / 255.0
        frames = self.transform(frames).permute(1, 0, 2, 3)
        return frames


def tensor_to_uint8(tensor):
    tensor = torch.clamp(tensor, -1.0, 1.0)
    tensor = (tensor + 1.0) / 2.0  # -1,1 -> 0,1; c,h,w
    tensor = (tensor.cpu().numpy() * 255).astype(np.uint8)
    return tensor


def main():
    def str2bool(v):
        if isinstance(v, bool):
            return v
        if v.lower() in ("yes", "true", "t", "y", "1"):
            return True
        elif v.lower() in ("no", "false", "f", "n", "0"):
            return False
        else:
            raise argparse.ArgumentTypeError("Boolean value expected.")

    parser = argparse.ArgumentParser()

    parser.add_argument(
        "--seed",
        type=int,
        default=42,
        help="the seed (for reproducible sampling)",
    )
    parser.add_argument(
        "--precision", type=str, help="evaluate at this precision", choices=["full", "autocast"], default="full"
    )
    parser.add_argument(
        "--config",
        type=str,
        default="configs/vidtok_kl_causal_488_4chn.yaml",
        help="path to config which constructs model",
    )
    parser.add_argument(
        "--ckpt",
        type=str,
        default="checkpoints/vidtok_kl_causal_488_4chn.ckpt",
        help="path to checkpoint of model",
    )
    parser.add_argument(
        "--output_video_dir",
        type=str,
        default="tmp",
        help="path to save the outputs",
    )
    parser.add_argument(
        "--input_video_path",
        type=str,
        default="assets/example.mp4",
        help="path to the input video",
    )
    parser.add_argument(
        "--input_height",
        type=int,
        default=256,
        help="height of the input video",
    )
    parser.add_argument(
        "--input_width",
        type=int,
        default=256,
        help="width of the input video",
    )
    parser.add_argument(
        "--sample_fps",
        type=int,
        default=30,
        help="sample fps",
    )
    parser.add_argument(
        "--chunk_size",
        type=int,
        default=16,
        help="the size of a chunk - we split a long video into several chunks",
    )
    parser.add_argument(
        "--read_long_video",
        action='store_true'
    )
    parser.add_argument(
        "--pad_gen_frames",
        action="store_true",
        help="Used only in causal mode. If True, pad frames generated in the last batch, else replicate the first frame instead",
    )
    parser.add_argument(
        "--concate_input",
        type=str2bool,
        const=True,
        default=True,
        nargs="?",
        help="",
    )

    args = parser.parse_args()
    seed_everything(args.seed)

    print0(f"[bold red]\[scripts.inference_reconstruct][/bold red] Evaluating model {args.ckpt}")
    device = torch.device("cuda") if torch.cuda.is_available() else torch.device("cpu")
    precision_scope = autocast if args.precision == "autocast" else nullcontext
    config = OmegaConf.load(args.config)

    os.makedirs(args.output_video_dir, exist_ok=True)

    model = load_model_from_config(args.config, args.ckpt)
    model.to(device).eval()
    assert args.chunk_size % model.encoder.time_downsample_factor == 0

    if args.read_long_video:
        assert hasattr(model, 'use_tiling'), "Tiling inference is needed to conduct long video reconstruction."
        print(f"Using tiling inference to save memory usage...")
        model.use_tiling = True
        model.t_chunk_enc = args.chunk_size
        model.t_chunk_dec = model.t_chunk_enc // model.encoder.time_downsample_factor
        model.use_overlap = True
        
    dataset = SingleVideoDataset(
        video_path=args.input_video_path, 
        input_height=args.input_height, 
        input_width=args.input_width, 
        sample_fps=args.sample_fps,
        chunk_size=args.chunk_size, 
        is_causal=model.is_causal,
        read_long_video=args.read_long_video
    )
    dataloader = torch.utils.data.DataLoader(dataset, batch_size=1, shuffle=False)

    inputs = []
    outputs = []
    with torch.no_grad(), precision_scope("cuda"):
        tic = time.time()
        for i, input in tqdm(enumerate(dataloader)):
            input = input.to(device)
            
            if model.is_causal and not args.read_long_video and args.pad_gen_frames:
                if i == 0:
                    _, xrec, _ = model(input)
                else:
                    _, xrec, _ = model(torch.cat([last_gen_frames, input], dim=2))
                xrec = xrec[:, :, -input.shape[2]:].clamp(-1, 1)
                last_gen_frames = xrec[:, :, (1 - model.encoder.time_downsample_factor):, :, :]
            else:
                _, xrec, _ = model(input)
                
            input = rearrange(input, "b c t h w -> (b t) c h w")
            inputs.append(input)
            xrec = rearrange(xrec.clamp(-1, 1), "b c t h w -> (b t) c h w")
            outputs.append(xrec)

        toc = time.time()

    # save the outputs as videos
    inputs = tensor_to_uint8(torch.cat(inputs, dim=0))
    inputs = rearrange(inputs, "t c h w -> t h w c")
    outputs = tensor_to_uint8(torch.cat(outputs, dim=0))
    outputs = rearrange(outputs, "t c h w -> t h w c")
    min_len = min(inputs.shape[0], outputs.shape[0])
    final = np.concatenate([inputs[:min_len], outputs[:min_len]], axis=2) if args.concate_input else outputs[:min_len]

    output_video_path = os.path.join(args.output_video_dir, f"{Path(args.input_video_path).stem}_reconstructed.mp4")
    write_video(output_video_path, final, args.sample_fps)

    print0(f"[bold red]Results saved in: {output_video_path}[/bold red]")
    print0(f"[bold red]\[scripts.inference_reconstruct][/bold red] Time taken: {toc - tic:.2f}s")


if __name__ == "__main__":
    main()