PerrinLin commited on
Commit
726435f
·
verified ·
1 Parent(s): 99f6167

HandDiffuse 3aaa125c73ea3ecc batch video-cd50f80037805bfd20a8

Browse files
episodes/tutting_12/cameras/primary_17/18.mp4 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:47b1afe4c47748e944cc3f39275510186f7845471b9e9a904f3d7b323cce7ff7
3
+ size 142093280
episodes/tutting_12/cameras/primary_17/18.mp4.privacy.json ADDED
@@ -0,0 +1,159 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "schema_version": 1,
3
+ "video_file": "18.mp4",
4
+ "source_sha256": "4aac9a45eb69710c5765e0fcef5d6f32015797865775bfef25fb43308b18ccb3",
5
+ "source_bytes": 156667256,
6
+ "output_sha256": "47b1afe4c47748e944cc3f39275510186f7845471b9e9a904f3d7b323cce7ff7",
7
+ "output_bytes": 142093280,
8
+ "processing": {
9
+ "detectors": [
10
+ "YuNet_2023mar_.65",
11
+ "CenterFace_.4",
12
+ "sideways_90deg_supplement_.65_stride30"
13
+ ],
14
+ "detection_stride": 5,
15
+ "interpolation": "linear bounding-box locations and sizes; every output frame masked",
16
+ "point_fields": [
17
+ "time_seconds",
18
+ "x",
19
+ "y",
20
+ "width",
21
+ "height"
22
+ ],
23
+ "boundary_policy": "All dual-detector tracks retained with sparse interpolation and 0.5s boundary extension; residual correction masks are strictly additive. No per-video human review claimed.",
24
+ "tracks": [
25
+ {
26
+ "width": 516,
27
+ "height": 796,
28
+ "start": 11.011511511511511,
29
+ "end": 12.094928261594928,
30
+ "points": [
31
+ [
32
+ 11.011511511511511,
33
+ 1875.3851623535156,
34
+ 1364,
35
+ 516,
36
+ 796
37
+ ],
38
+ [
39
+ 11.511511511511511,
40
+ 1875.3851623535156,
41
+ 1364,
42
+ 516,
43
+ 796
44
+ ],
45
+ [
46
+ 11.594928261594928,
47
+ 1845.9229736328125,
48
+ 1372,
49
+ 506,
50
+ 788
51
+ ],
52
+ [
53
+ 12.094928261594928,
54
+ 1845.9229736328125,
55
+ 1372,
56
+ 506,
57
+ 788
58
+ ]
59
+ ],
60
+ "detections": 2
61
+ },
62
+ {
63
+ "width": 118,
64
+ "height": 224,
65
+ "start": 47.080914247580914,
66
+ "end": 48.080914247580914,
67
+ "points": [
68
+ [
69
+ 47.080914247580914,
70
+ 1646.8403015136719,
71
+ 927.3199615478516,
72
+ 118,
73
+ 224
74
+ ],
75
+ [
76
+ 47.580914247580914,
77
+ 1646.8403015136719,
78
+ 927.3199615478516,
79
+ 118,
80
+ 224
81
+ ],
82
+ [
83
+ 48.080914247580914,
84
+ 1646.8403015136719,
85
+ 927.3199615478516,
86
+ 118,
87
+ 224
88
+ ]
89
+ ],
90
+ "detections": 1
91
+ }
92
+ ],
93
+ "face_track_count": 2,
94
+ "codec": "h264",
95
+ "encoder": "h264_nvenc",
96
+ "cq": 23,
97
+ "pixelation_grid": [
98
+ 8,
99
+ 8
100
+ ],
101
+ "algorithm": "mano20-alpha-sparse5-v1",
102
+ "residual_correction_policy": "Additional masks only; original detector-track masks retained.",
103
+ "hand_protection": {
104
+ "outward_padding_pixels": 20,
105
+ "region": "separate convex hulls of recorded projected MANO vertices",
106
+ "compositing": "opaque original-source alpha inside protection, face mosaic outside",
107
+ "model_or_camera_refit": false,
108
+ "missing_motion_frame_indices": [],
109
+ "missing_pose_policy": "preserve original pixels for entire frame"
110
+ },
111
+ "projection": {
112
+ "episode": "tutting_12",
113
+ "camera_uid": "primary_17",
114
+ "parameter_namespace": "primary_original",
115
+ "original_parameter_id": "17",
116
+ "calibration_sha256": "2a7b368805838ea1d2212685c4e878cae1044a6d9540aa62475aa6192b9b5029",
117
+ "source_frame_map_sha256": "238785be3087141b5a95d3c899a8223f5989a49c201b6f16542a7b2f7a84adeb",
118
+ "models": {
119
+ "left": {
120
+ "motion_sha256": "3ade4222aed624cadbf290c78bf13e19387808346c68a87517190a153ded3c44",
121
+ "model_sha256": "c4022f7083f2ca7c78b2b3d595abbab52debd32b09d372b16923a801f0ea6a30"
122
+ },
123
+ "right": {
124
+ "motion_sha256": "aa3e537a9b97fc357a2a52e06608ffd01626cbb29aebd6d72fb744e40dc55553",
125
+ "model_sha256": "45d60aa3b27ef9107a7afd4e00808f307fd91111e1cfa35afd5c4a62de264767"
126
+ }
127
+ },
128
+ "new_camera_fit": false,
129
+ "convention": "V_source=V_raw-J_raw[0]+saved_root; X_camera=R*X_source+t; p=K*X_camera",
130
+ "approximate_mapping_policy": "union of recorded selected and alternative hand projections",
131
+ "projection_scope": "Consume existing mapping, no new frame-map integrity audit."
132
+ }
133
+ },
134
+ "validation": {
135
+ "status": "PASSED_HAND_PRIORITY_POLICY",
136
+ "frames": 3665,
137
+ "timestamp_equality": true,
138
+ "dimensions_preserved": true,
139
+ "audio_streams": 0,
140
+ "residual_check_frames": 733,
141
+ "elapsed_seconds": 48.394435063994024,
142
+ "scope": "Sparse dual-detector residual check, not a proof against all detector blind spots.",
143
+ "allowed_hand_protection_residuals": []
144
+ },
145
+ "visual_review": {
146
+ "mosaic_visible": true,
147
+ "hands_reviewed": false,
148
+ "false_positive_residual_count": 0,
149
+ "scope": "Automatic exact frame/PTS/dimension/audio checks and offset sparse dual-face detection; approved MANO20 alpha compositor. No per-video human visual review claimed.",
150
+ "kind": "automated_policy_checks"
151
+ },
152
+ "dimensions": [
153
+ 3840,
154
+ 2160
155
+ ],
156
+ "frames": 3665,
157
+ "source_time_base": "1/11988",
158
+ "timestamps_sha256": "629acd8b9b6707f25cd771a663d44b608b1bdba2998343e511a02d359a96862d"
159
+ }