File size: 6,288 Bytes
da3ed5b
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
"""
Visualize GT depth - both reverse and non-reverse versions
Must match exactly with visualize_depth.py sample selection
"""
import os
import sys
import numpy as np
from PIL import Image
import matplotlib.pyplot as plt
import matplotlib
matplotlib.use('Agg')
from tqdm import tqdm

# Dataset paths - MUST match visualize_depth.py
KITTI_BASE = '/home/ywan0794/datasets/eval/moge_style_eval/KITTI'
DDAD_BASE = '/home/ywan0794/datasets/eval/moge_style_eval/DDAD/val'
OUTPUT_DIR = '/home/ywan0794/MoGe/vis_output'


def colorize_depth(depth, cmap='Spectral', reverse=False, mask_invalid=False):
    depth = depth.copy()
    if mask_invalid:
        invalid_mask = depth <= 0
        valid_depth = depth[~invalid_mask]
        if len(valid_depth) > 0:
            vmin = np.percentile(valid_depth, 2)
            vmax = np.percentile(valid_depth, 98)
        else:
            vmin, vmax = 0, 1
    else:
        vmin = np.percentile(depth, 2)
        vmax = np.percentile(depth, 98)

    depth = (depth - vmin) / (vmax - vmin + 1e-8)
    depth = np.clip(depth, 0, 1)

    if reverse:
        depth = 1 - depth

    cm = plt.get_cmap(cmap)
    colored = cm(depth)[:, :, :3]
    colored = (colored * 255).astype(np.uint8)

    if mask_invalid:
        colored[invalid_mask] = 0

    return colored


def load_gt_depth(depth_path):
    depth = np.array(Image.open(depth_path))
    if depth.dtype == np.uint16:
        depth = depth.astype(np.float32) / 256.0
    elif depth.dtype == np.uint8:
        depth = depth.astype(np.float32)
    return depth


def main():
    print("=" * 50, flush=True)
    print("GT Depth Visualization (both versions)", flush=True)
    print("=" * 50, flush=True)

    # Create output directories
    for dataset in ['KITTI', 'DDAD']:
        os.makedirs(os.path.join(OUTPUT_DIR, dataset, 'gt'), exist_ok=True)
        os.makedirs(os.path.join(OUTPUT_DIR, dataset, 'gt_reverse'), exist_ok=True)

    # ============================================
    # KITTI - MUST match visualize_depth.py exactly
    # ============================================
    print("\nCollecting KITTI samples...", flush=True)
    kitti_samples = []
    for drive in os.listdir(KITTI_BASE):
        drive_path = os.path.join(KITTI_BASE, drive, 'image_02')
        if os.path.isdir(drive_path):
            for frame in sorted(os.listdir(drive_path)):
                sample_dir = os.path.join(drive_path, frame)
                img_path = os.path.join(sample_dir, 'image.jpg')
                gt_path = os.path.join(sample_dir, 'depth.png')
                # MUST check both img and gt exist - same as visualize_depth.py
                if os.path.exists(img_path) and os.path.exists(gt_path):
                    kitti_samples.append({
                        'image': img_path,
                        'gt': gt_path,
                        'name': f"{drive}_{frame}"
                    })

    # Same random seed and selection as visualize_depth.py
    np.random.seed(42)
    num_samples = 10  # Must match --num-samples in visualize_depth.py
    kitti_selected = np.random.choice(len(kitti_samples), min(num_samples, len(kitti_samples)), replace=False)

    print(f"Processing {len(kitti_selected)} KITTI GT samples...", flush=True)
    for idx, i in tqdm(enumerate(kitti_selected), total=len(kitti_selected), desc="KITTI GT", file=sys.stdout):
        sample = kitti_samples[i]
        gt_depth = load_gt_depth(sample['gt'])

        # Non-reverse version
        gt_colored = colorize_depth(gt_depth, reverse=False, mask_invalid=True)
        Image.fromarray(gt_colored).save(os.path.join(OUTPUT_DIR, 'KITTI', 'gt', f"{idx:03d}.png"))

        # Reverse version
        gt_colored_rev = colorize_depth(gt_depth, reverse=True, mask_invalid=True)
        Image.fromarray(gt_colored_rev).save(os.path.join(OUTPUT_DIR, 'KITTI', 'gt_reverse', f"{idx:03d}.png"))

    # ============================================
    # DDAD - MUST match visualize_depth.py exactly
    # ============================================
    print("\nCollecting DDAD samples...", flush=True)
    ddad_samples = []
    for scene in sorted(os.listdir(DDAD_BASE)):
        scene_path = os.path.join(DDAD_BASE, scene)
        if os.path.isdir(scene_path):
            for cam in sorted(os.listdir(scene_path)):
                sample_dir = os.path.join(scene_path, cam)
                img_path = os.path.join(sample_dir, 'image.jpg')
                gt_path = os.path.join(sample_dir, 'depth.png')
                # MUST check both img and gt exist - same as visualize_depth.py
                if os.path.exists(img_path) and os.path.exists(gt_path):
                    ddad_samples.append({
                        'image': img_path,
                        'gt': gt_path,
                        'name': f"{scene}_{cam}"
                    })

    # Same random seed and selection as visualize_depth.py
    # Note: seed was already set to 42 above, and kitti_selected consumed some random numbers
    # We need to match the exact sequence
    ddad_selected = np.random.choice(len(ddad_samples), min(num_samples, len(ddad_samples)), replace=False)

    print(f"Processing {len(ddad_selected)} DDAD GT samples...", flush=True)
    for idx, i in tqdm(enumerate(ddad_selected), total=len(ddad_selected), desc="DDAD GT", file=sys.stdout):
        sample = ddad_samples[i]
        gt_depth = load_gt_depth(sample['gt'])

        # Non-reverse version
        gt_colored = colorize_depth(gt_depth, reverse=False, mask_invalid=True)
        Image.fromarray(gt_colored).save(os.path.join(OUTPUT_DIR, 'DDAD', 'gt', f"{idx:03d}.png"))

        # Reverse version
        gt_colored_rev = colorize_depth(gt_depth, reverse=True, mask_invalid=True)
        Image.fromarray(gt_colored_rev).save(os.path.join(OUTPUT_DIR, 'DDAD', 'gt_reverse', f"{idx:03d}.png"))

    print("\n" + "=" * 50, flush=True)
    print("Done!", flush=True)
    print("Output:", flush=True)
    print(f"  {OUTPUT_DIR}/KITTI/gt/         (non-reverse)", flush=True)
    print(f"  {OUTPUT_DIR}/KITTI/gt_reverse/ (reverse)", flush=True)
    print(f"  {OUTPUT_DIR}/DDAD/gt/          (non-reverse)", flush=True)
    print(f"  {OUTPUT_DIR}/DDAD/gt_reverse/  (reverse)", flush=True)
    print("=" * 50, flush=True)


if __name__ == '__main__':
    main()