Dataset Viewer
Auto-converted to Parquet Duplicate
Search is not available for this dataset
video
video
label
class label
1hdepic_P01-20240202-110250_0001710_A
2hdepic_P01-20240202-110250_0001710_AVBearly
3hdepic_P01-20240202-110250_0001710_TCC
4hdepic_P01-20240202-110250_0001710_TCR
5hdepic_P01-20240202-110250_0001710_Vhand
6hdepic_P01-20240202-110250_0001710_Vnoun
7hdepic_P01-20240202-110250_0001710_mute
8hdepic_P01-20240202-110250_0001710_pass
9hdepic_P01-20240202-110250_0012785_A
10hdepic_P01-20240202-110250_0012785_AVBearly
11hdepic_P01-20240202-110250_0012785_AVBlate
12hdepic_P01-20240202-110250_0012785_TCC
13hdepic_P01-20240202-110250_0012785_TCR
14hdepic_P01-20240202-110250_0012785_Vhand
15hdepic_P01-20240202-110250_0012785_Vnoun
16hdepic_P01-20240202-110250_0012785_mute
17hdepic_P01-20240202-110250_0012785_pass
18hdepic_P01-20240202-110250_0021727_A
19hdepic_P01-20240202-110250_0021727_AVBearly
20hdepic_P01-20240202-110250_0021727_TCC
21hdepic_P01-20240202-110250_0021727_TCR
22hdepic_P01-20240202-110250_0021727_Vnoun
23hdepic_P01-20240202-110250_0021727_mute
24hdepic_P01-20240202-110250_0021727_pass
25hdepic_P01-20240202-110250_0025014_A
26hdepic_P01-20240202-110250_0025014_AVBearly
27hdepic_P01-20240202-110250_0025014_TCC
28hdepic_P01-20240202-110250_0025014_TCR
29hdepic_P01-20240202-110250_0025014_Vhand
30hdepic_P01-20240202-110250_0025014_Vnoun
31hdepic_P01-20240202-110250_0025014_mute
32hdepic_P01-20240202-110250_0025014_pass
33hdepic_P01-20240202-161354_0011891_A
34hdepic_P01-20240202-161354_0011891_AVBearly
35hdepic_P01-20240202-161354_0011891_AVBlate
36hdepic_P01-20240202-161354_0011891_TCC
37hdepic_P01-20240202-161354_0011891_TCR
38hdepic_P01-20240202-161354_0011891_Vhand
39hdepic_P01-20240202-161354_0011891_Vnoun
40hdepic_P01-20240202-161354_0011891_mute
41hdepic_P01-20240202-161354_0011891_pass
42hdepic_P01-20240202-161354_0025401_A
43hdepic_P01-20240202-161354_0025401_AVBearly
44hdepic_P01-20240202-161354_0025401_AVBlate
45hdepic_P01-20240202-161354_0025401_TCC
46hdepic_P01-20240202-161354_0025401_TCR
47hdepic_P01-20240202-161354_0025401_Vhand
48hdepic_P01-20240202-161354_0025401_Vnoun
49hdepic_P01-20240202-161354_0025401_mute
50hdepic_P01-20240202-161354_0025401_pass
51hdepic_P01-20240202-161948_0033531_A
52hdepic_P01-20240202-161948_0033531_AVBearly
53hdepic_P01-20240202-161948_0033531_TCC
54hdepic_P01-20240202-161948_0033531_TCR
55hdepic_P01-20240202-161948_0033531_Vhand
56hdepic_P01-20240202-161948_0033531_Vnoun
57hdepic_P01-20240202-161948_0033531_mute
58hdepic_P01-20240202-161948_0033531_pass
59hdepic_P01-20240202-161948_0046347_A
60hdepic_P01-20240202-161948_0046347_AVBearly
61hdepic_P01-20240202-161948_0046347_TCC
62hdepic_P01-20240202-161948_0046347_TCR
63hdepic_P01-20240202-161948_0046347_Vhand
64hdepic_P01-20240202-161948_0046347_Vnoun
65hdepic_P01-20240202-161948_0046347_mute
66hdepic_P01-20240202-161948_0046347_pass
67hdepic_P01-20240202-161948_0055835_A
68hdepic_P01-20240202-161948_0055835_AVBlate
69hdepic_P01-20240202-161948_0055835_TCC
70hdepic_P01-20240202-161948_0055835_TCR
71hdepic_P01-20240202-161948_0055835_Vnoun
72hdepic_P01-20240202-161948_0055835_mute
73hdepic_P01-20240202-161948_0055835_pass
74hdepic_P01-20240202-161948_0069559_A
75hdepic_P01-20240202-161948_0069559_AVBearly
76hdepic_P01-20240202-161948_0069559_AVBlate
77hdepic_P01-20240202-161948_0069559_TCC
78hdepic_P01-20240202-161948_0069559_TCR
79hdepic_P01-20240202-161948_0069559_Vhand
80hdepic_P01-20240202-161948_0069559_Vnoun
81hdepic_P01-20240202-161948_0069559_mute
82hdepic_P01-20240202-161948_0069559_pass
83hdepic_P01-20240202-171220_0010143_A
84hdepic_P01-20240202-171220_0010143_AVBearly
85hdepic_P01-20240202-171220_0010143_AVBlate
86hdepic_P01-20240202-171220_0010143_TCC
87hdepic_P01-20240202-171220_0010143_TCR
88hdepic_P01-20240202-171220_0010143_Vnoun
89hdepic_P01-20240202-171220_0010143_mute
90hdepic_P01-20240202-171220_0010143_pass
91hdepic_P01-20240202-171220_0034731_A
92hdepic_P01-20240202-171220_0034731_AVBearly
93hdepic_P01-20240202-171220_0034731_AVBlate
94hdepic_P01-20240202-171220_0034731_TCC
95hdepic_P01-20240202-171220_0034731_TCR
96hdepic_P01-20240202-171220_0034731_Vnoun
97hdepic_P01-20240202-171220_0034731_mute
98hdepic_P01-20240202-171220_0034731_pass
99hdepic_P01-20240202-175627_0010654_A
100hdepic_P01-20240202-175627_0010654_AVBlate
End of preview. Expand in Data Studio

HD-EPIC real-video benchmark for T2AV judge evaluation

Clips are cut from HD-EPIC (Perrett et al., "HD-EPIC: A Highly-Detailed Egocentric Video Dataset", CVPR 2025). For research use, follow the original HD-EPIC license and cite the paper.

What this is

Real kitchen clips used as "correct" videos. Prompts are written from the HD-EPIC narration and sound annotations, then one axis is broken on purpose to make FAIL items. Each item = one clip + one prompt + a GT label per axis (A, V, AVB, TC-Rel, TC-Cnt). CC is not scored on real videos.

194 original windows, 1,523 items in total:

Variant Folder suffix Broken axis n How it is made
original _pass none (all PASS) 194 original clip and prompt
mute _mute A 194 audio removed
absent-sound claim _A A 194 prompt claims a sound that is not there
hand swap _Vhand V 134 prompt and event list use the other hand
noun swap _Vnoun V 194 prompt and event list use another object
audio shift early _AVBearly AVB 124 audio track shifted earlier
audio shift late _AVBlate AVB 101 audio track shifted later
order swap _TCR TC-Rel 194 order of the two events reversed in the prompt
count off _TCC TC-Cnt 194 required count changed by +1

Variants that only change the prompt (A, Vhand, Vnoun, TCR, TCC) reuse the original clip, so the same video appears in several folders.

GT hierarchy

GT follows the same hierarchy the judge uses, at event level:

  • A FAIL on an event → AVB FAIL for that event
  • V FAIL → AVB FAIL only if the broken event is the sounding anchor (36 of 134 hand swaps, 0 noun swaps)
  • A, V or AVB FAIL → TC-Rel and TC-Cnt FAIL

broken_event_idx / broken_event_kind columns in the GT csv tell which event was changed.

Files

  • videos/<item_id>/<item_id>_v1_raw.mp4 : one clip per item (1,523), same layout as the judge input
  • hdepic_medium_prompts.json : prompts in the new_eval.py input format
  • hdepic_medium_gt.csv : per-axis GT (hierarchical, use this one)
  • items_hdepic_medium_with_clips_v2.json : source items with events, narration timestamps and variant info

For AVB tools

Audio shift items (225) are the real AVB negatives. Originals (194) are the positives for false alarm rate. The Gemini judge (single agent, no tools) detects early shifts 13% and late shifts 32% of the time, so this is where an AVB tool can show value.

Downloads last month
73