Parallelized Autoregressive Decoding for Omni-Modal Dense Video Captioning
Paper โข 2607.02963 โข Published โข 27
Dataset Preview
The full dataset viewer is not available (click to read why). Only showing a preview of the rows.
The dataset generation failed
Error code: DatasetGenerationError
Exception: CastError
Message: Couldn't cast
id: string
duration: double
category: string
video: string
use_audio: bool
task: string
longvale_task: string
conversations: list<item: struct<from: string, value: string>>
child 0, item: struct<from: string, value: string>
child 0, from: string
child 1, value: string
tgt: list<item: double>
child 0, item: double
src: list<item: double>
child 0, item: double
video_id: string
gpt_value_segments: list<item: string>
child 0, item: string
to
{'id': Value('string'), 'duration': Value('float64'), 'category': Value('string'), 'video': Value('string'), 'task': Value('string'), 'longvale_task': Value('string'), 'use_audio': Value('bool'), 'conversations': List({'from': Value('string'), 'value': Value('string')}), 'tgt': List(Value('float64')), 'src': List(Value('float64')), 'gpt_value_segments': List(Value('string'))}
because column names don't match
Traceback: Traceback (most recent call last):
File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1827, in _prepare_split_single
for key, table in generator:
^^^^^^^^^
File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 613, in wrapped
for item in generator(*args, **kwargs):
~~~~~~~~~^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 343, in _generate_tables
self._cast_table(pa_table, json_field_paths=json_field_paths),
~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 132, in _cast_table
pa_table = table_cast(pa_table, self.info.features.arrow_schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2378, in table_cast
return cast_table_to_schema(table, schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2306, in cast_table_to_schema
raise CastError(
...<3 lines>...
)
datasets.table.CastError: Couldn't cast
id: string
duration: double
category: string
video: string
use_audio: bool
task: string
longvale_task: string
conversations: list<item: struct<from: string, value: string>>
child 0, item: struct<from: string, value: string>
child 0, from: string
child 1, value: string
tgt: list<item: double>
child 0, item: double
src: list<item: double>
child 0, item: double
video_id: string
gpt_value_segments: list<item: string>
child 0, item: string
to
{'id': Value('string'), 'duration': Value('float64'), 'category': Value('string'), 'video': Value('string'), 'task': Value('string'), 'longvale_task': Value('string'), 'use_audio': Value('bool'), 'conversations': List({'from': Value('string'), 'value': Value('string')}), 'tgt': List(Value('float64')), 'src': List(Value('float64')), 'gpt_value_segments': List(Value('string'))}
because column names don't match
The above exception was the direct cause of the following exception:
Traceback (most recent call last):
File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 1369, in compute_config_parquet_and_info_response
parquet_operations, partial, estimated_dataset_info = stream_convert_to_parquet(
~~~~~~~~~~~~~~~~~~~~~~~~~^
builder, max_dataset_size_bytes=max_dataset_size_bytes
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
)
^
File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 948, in stream_convert_to_parquet
builder._prepare_split(split_generator=splits_generators[split], file_format="parquet")
~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1694, in _prepare_split
for job_id, done, content in self._prepare_split_single(
~~~~~~~~~~~~~~~~~~~~~~~~~~^
gen_kwargs=gen_kwargs, job_id=job_id, **_prepare_split_args
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
):
^
File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1880, in _prepare_split_single
raise DatasetGenerationError("An error occurred while generating the dataset") from e
datasets.exceptions.DatasetGenerationError: An error occurred while generating the datasetNeed help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
id string | duration float64 | category string | video string | task string | longvale_task string | use_audio bool | conversations list | tgt list | src list | gpt_value_segments list |
|---|---|---|---|---|---|---|---|---|---|---|
FIObpRGDBY4 | 105 | News & Politics | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/FIObpRGDBY4.mp4 | v2t | grounding | true | [
{
"from": "human",
"value": "<video>\nLocalize the visual content described by the given textual query < An outdoor sports field featuring several dark, curved training barriers on the grass, a tall metal goalpost, and a large building with a white roof and dark walls in the background. A few people are vis... | [
40.24,
42.976
] | [] | null |
FIObpRGDBY4 | 105 | News & Politics | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/FIObpRGDBY4.mp4 | a2t | grounding | true | [
{
"from": "human",
"value": "<video>\nFind the video segment that corresponds to the given textual query < A man speaks: \"Now, Friday's game with Gloucester has been cancelled.\" > and determine its start and end seconds."
},
{
"from": "gpt",
"value": "<G>"
}
] | [
40.24,
42.976
] | [] | null |
FIObpRGDBY4 | 105 | News & Politics | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/FIObpRGDBY4.mp4 | t2v | seg_captioning | true | [
{
"from": "human",
"value": "<video>\nTell me about the visual events from <G> in the video."
},
{
"from": "gpt",
"value": "<G>A news reporter stands on a sports field with stadium seating and goalposts in the background, under twilight conditions. On-screen graphics identify the reporter as Jim... | [
91.791,
104.304
] | [
91.791,
104.304
] | null |
FIObpRGDBY4 | 105 | News & Politics | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/FIObpRGDBY4.mp4 | t2a | seg_captioning | true | [
{
"from": "human",
"value": "<video>\nTell me about the audio events from <G> in the video."
},
{
"from": "gpt",
"value": "<G>A male voice speaks, \"So once again, Friday's game against Gloucester is off, and the immediate future of the Winthrop High School football program is uncertain, pending... | [
91.791,
104.304
] | [
91.791,
104.304
] | null |
FIObpRGDBY4 | 105 | News & Politics | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/FIObpRGDBY4.mp4 | dvc | captioning | true | [
{
"from": "human",
"value": "<video>\nCan you go through the video and describe what took place at different time intervals?"
},
{
"from": "gpt",
"value": "<G><G><G><G><G><G><G><G><G><G><G><shift><G><G><G><G><G><G><G><G><G><G><G><shift>Visual: A female news anchor, Lisa Hughes, sits at a news de... | [
0,
12.178,
12.212,
16.449,
16.483,
24.024,
24.057,
27.494,
27.527,
37.47,
37.504,
40.206,
40.24,
42.976,
43.009,
46.679,
46.713,
62.796,
62.829,
91.758,
91.791,
104.304,
0,
12.178,
12.212,
16.449,
16.483,
24.024,
24.057,
27.494,
27.527,
37.47,
37.504,
... | [] | [
"Visual: A female news anchor, Lisa Hughes, sits at a news desk with a screen behind her displaying \"FOOTBALL SCANDAL\" over a football field graphic. Another screen behind her shows a blurry night cityscape. A lower-third graphic identifies her and the news channel. Audio: A woman speaks, \"A high school football... |
FIObpRGDBY4 | 105 | News & Politics | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/FIObpRGDBY4.mp4 | v2a | captioning | true | [
{
"from": "human",
"value": "<video>\nGiven the visual content < A female news anchor, Lisa Hughes, sits at a news desk with a screen behind her displaying \"FOOTBALL SCANDAL\" over a football field graphic. Another screen behind her shows a blurry night cityscape. A lower-third graphic identifies her and t... | [
0,
12.178
] | [] | null |
FIObpRGDBY4 | 105 | News & Politics | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/FIObpRGDBY4.mp4 | a2v | captioning | true | [
{
"from": "human",
"value": "<video>\nFind the video segment that corresponds to the given textual query < A woman speaks, \"A high school football player is suspended and could face charges for videotaping a teammate in the locker room. And as Jim Smith shows us from Winthrop, there could be a bigger lesso... | [
0,
12.178
] | [] | null |
bBIk8Hmnl9s | 77 | News & Politics | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/bBIk8Hmnl9s.mp4 | v2t | grounding | true | [
{
"from": "human",
"value": "<video>\nBetween which frames can we find: < Three women are seated at a table on a talk show set, with mugs and papers visible. One woman wears a light blue jacket, another a pink dress, and the third a floral dress. News-related text overlays appear on the screen. >?"
},
{... | [
50.417,
52.352
] | [] | null |
bBIk8Hmnl9s | 77 | News & Politics | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/bBIk8Hmnl9s.mp4 | a2t | grounding | true | [
{
"from": "human",
"value": "<video>\nLocalize the audio content described by the given textual query < A male voice says, \"According to a published report...\" > in the video, and output the start and end timestamps in seconds."
},
{
"from": "gpt",
"value": "<G>"
}
] | [
50.417,
52.352
] | [] | null |
bBIk8Hmnl9s | 77 | News & Politics | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/bBIk8Hmnl9s.mp4 | t2v | seg_captioning | true | [
{
"from": "human",
"value": "<video>\nFocus on the video event in <G> and describe what is happening from the visual perspective."
},
{
"from": "gpt",
"value": "<G>Hillary Clinton wearing a teal jacket and earrings in a television studio setting for \"The View.\""
}
] | [
24.924,
26.126
] | [
24.924,
26.126
] | null |
bBIk8Hmnl9s | 77 | News & Politics | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/bBIk8Hmnl9s.mp4 | t2a | seg_captioning | true | [
{
"from": "human",
"value": "<video>\nOutput the audio information of the audio segment from <G>."
},
{
"from": "gpt",
"value": "<G>A male voice says, \"is on the panel.\""
}
] | [
24.924,
26.126
] | [
24.924,
26.126
] | null |
bBIk8Hmnl9s | 77 | News & Politics | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/bBIk8Hmnl9s.mp4 | dvc | captioning | true | [
{
"from": "human",
"value": "<video>\nLocalize a series of activity events in the video, output the start and end timestamp for each event, and <time> - <time>, <event>. Describe each event with sentences. The output format of each predicted event should be like: โstart - <time> - <time>, <event>. end secon... | [
0,
5.572,
5.605,
9.843,
9.876,
18.418,
18.451,
22.422,
22.455,
23.656,
23.69,
24.891,
24.924,
26.126,
26.159,
28.762,
28.795,
29.829,
29.863,
35.101,
35.135,
37.203,
37.237,
38.405,
38.438,
40.04,
40.073,
42.008,
42.042,
45.745,
45.779,
47.547,
47.58,
... | [] | [
"Visual: A woman with long brown hair speaks in a close-up, then appears as part of a panel of five women seated at a table on a television studio set. One panelist wears a \"BIRTHDAY GIRL\" sash and tiara, and another panelist gives a high-five. Audio: A woman speaks, \"This is my last day at The View.\" An audien... |
bBIk8Hmnl9s | 77 | News & Politics | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/bBIk8Hmnl9s.mp4 | v2a | captioning | true | [
{
"from": "human",
"value": "<video>\nFind the video segment that corresponds to the given textual query < A woman with brown hair, wearing a patterned top, speaking in a television studio setting with floral decorations in the background. Text overlays appear at the bottom of the screen. > and tell me what... | [
47.58,
48.748
] | [] | null |
bBIk8Hmnl9s | 77 | News & Politics | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/bBIk8Hmnl9s.mp4 | a2v | captioning | true | [
{
"from": "human",
"value": "<video>\nWhen the audio indicates < A male voice speaks, \"who were upset,\" which is followed by a clicking sound. >, what visual scene or actions can be observed in the video? Describe the visual content."
},
{
"from": "gpt",
"value": "<G>A woman with brown hair, w... | [
47.58,
48.748
] | [] | null |
s8imRcg9tHk | 180 | Sports | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/s8imRcg9tHk.mp4 | v2t | grounding | true | [
{
"from": "human",
"value": "<video>\nLocalize the visual content described by the given textual query < Female athletes in red and black uniforms run on an indoor track during a competition, with spectators visible in the background stands. Broadcast graphics are superimposed on the lower part of the scree... | [
161.995,
168.067
] | [] | null |
s8imRcg9tHk | 180 | Sports | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/s8imRcg9tHk.mp4 | a2t | grounding | true | [
{
"from": "human",
"value": "<video>\nOutput the start and end time of the video segment corresponding to this audio information: < A female voice speaks, \"Junior Ashley Kelly broke the school record in the 400 meter with a time of 53.\" >."
},
{
"from": "gpt",
"value": "<G>"
}
] | [
161.995,
168.067
] | [] | null |
s8imRcg9tHk | 180 | Sports | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/s8imRcg9tHk.mp4 | t2v | seg_captioning | true | [
{
"from": "human",
"value": "<video>\nWhat transpired from <G> in the video, according to the visual information?"
},
{
"from": "gpt",
"value": "<G>Athletes compete in an indoor track and field event, running on a track past spectators in the stands."
}
] | [
148.715,
151.618
] | [
148.715,
151.618
] | null |
s8imRcg9tHk | 180 | Sports | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/s8imRcg9tHk.mp4 | t2a | seg_captioning | true | [
{
"from": "human",
"value": "<video>\nOutput the audio information of the audio segment from <G>."
},
{
"from": "gpt",
"value": "<G>A female voice speaks, \"in the 200-meter dash while sophomore,\" accompanied by rock music and the sounds of a crowd."
}
] | [
148.715,
151.618
] | [
148.715,
151.618
] | null |
s8imRcg9tHk | 180 | Sports | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/s8imRcg9tHk.mp4 | dvc | captioning | true | [
{
"from": "human",
"value": "<video>\nDetect and report the start and end timestamps of activity events in the video, along with descriptions."
},
{
"from": "gpt",
"value": "<G><G><G><G><G><G><G><G><G><G><G><G><G><G><G><G><G><G><G><G><G><G><G><G><G><shift><G><G><G><G><G><G><G><G><G><G><G><G><G><... | [
0,
9.576,
9.609,
12.779,
12.812,
19.686,
19.719,
28.995,
29.029,
46.046,
46.079,
51.518,
51.551,
60.46,
60.493,
74.441,
74.474,
76.743,
76.776,
80.346,
80.38,
90.023,
90.056,
97.397,
97.43,
104.537,
104.571,
114.681,
114.714,
128.895,
128.928,
134.734,
... | [] | [
"Visual: An animated title sequence for \"Fighting Illini Insider Review\" transitions to a wrestling match between two male athletes on a mat, with a referee and spectators visible in the background. Audio: Electronic music plays. A woman speaks, \"The Fighting Illini wrestling team took on another Big Ten opponen... |
s8imRcg9tHk | 180 | Sports | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/s8imRcg9tHk.mp4 | v2a | captioning | true | [
{
"from": "human",
"value": "<video>\nFind the video segment that corresponds to the given textual query < Tennis players hitting and serving tennis balls on an indoor court, followed by players shaking hands over the net. > and tell me what was going from audio perspective."
},
{
"from": "gpt",
... | [
29.029,
46.046
] | [] | null |
s8imRcg9tHk | 180 | Sports | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/s8imRcg9tHk.mp4 | a2v | captioning | true | [
{
"from": "human",
"value": "<video>\nBased on the given audio cue < A female voice speaks, \"The men's tennis team lost to number three Tennessee 5-2 in Champagne this weekend. However, number 11 ranked Denis Nevolo defeated the number one opponent in the country. Nevolo earned his 18th win of the year and... | [
29.029,
46.046
] | [] | null |
6UQfcckT2oU | 331 | Autos & Vehicles | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/6UQfcckT2oU.mp4 | v2t | grounding | true | [
{
"from": "human",
"value": "<video>\nDetect and report the start and end timestamps of the video segment that semantically matches the given textual query < A person is wet sanding the fender of a black car, which has bright green wheels, outdoors. >."
},
{
"from": "gpt",
"value": "<G>"
}
] | [
194.794,
201.334
] | [] | null |
6UQfcckT2oU | 331 | Autos & Vehicles | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/6UQfcckT2oU.mp4 | a2t | grounding | true | [
{
"from": "human",
"value": "<video>\nLocalize the audio content described by the given textual query < Instrumental hip-hop music plays, featuring a steady beat with cymbals and a bassline. > in the video, and output the start and end timestamps in seconds."
},
{
"from": "gpt",
"value": "<G>"
... | [
194.794,
201.334
] | [] | null |
6UQfcckT2oU | 331 | Autos & Vehicles | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/6UQfcckT2oU.mp4 | t2v | seg_captioning | true | [
{
"from": "human",
"value": "<video>\nProvide details about the visual events from <G> in the video."
},
{
"from": "gpt",
"value": "<G>A person applies a scratch repair pen to cover scratches on the black fender of a car with green rims."
}
] | [
27.627,
65.365
] | [
27.627,
65.365
] | null |
6UQfcckT2oU | 331 | Autos & Vehicles | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/6UQfcckT2oU.mp4 | t2a | seg_captioning | true | [
{
"from": "human",
"value": "<video>\nWhat transpired from <G> in the video, according to the audio information?"
},
{
"from": "gpt",
"value": "<G>A male voice speaks, \"Okay, so the first thing that's telling me about is that this tip removes the rust and the loose paint. So I'm just going to..... | [
27.627,
65.365
] | [
27.627,
65.365
] | null |
6UQfcckT2oU | 331 | Autos & Vehicles | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/6UQfcckT2oU.mp4 | dvc | captioning | true | [
{
"from": "human",
"value": "<video>\nI'd like to know what events transpired during specific time intervals in the video. Could you please elaborate?"
},
{
"from": "gpt",
"value": "<G><G><G><G><G><G><G><G><G><G><G><G><G><G><G><shift><G><G><G><G><G><G><G><G><G><G><G><G><G><G><G><shift>Visual: A ... | [
0,
27.594,
27.627,
65.365,
65.398,
123.056,
123.089,
142.242,
142.275,
171.171,
171.204,
178.144,
178.178,
184.817,
184.851,
194.761,
194.794,
201.334,
201.367,
236.336,
236.369,
279.545,
279.579,
282.348,
282.382,
285.318,
285.351,
319.152,
319.185,
330.93,... | [] | [
"Visual: A person's hand holds a scratch repair product, then points to and displays prominent scratches on the black fender of a car with bright green rims. Audio: A person speaks: \"Alright guys, usually I'm not much for cosmetics, but I did want this car to be clean. And this scratch has been bothering me since ... |
6UQfcckT2oU | 331 | Autos & Vehicles | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/6UQfcckT2oU.mp4 | v2a | captioning | true | [
{
"from": "human",
"value": "<video>\nWhat sounds or audio events accompany the visual scene described as < A black sports car with bright green wheels is detailed, transitioning its surface from a dull appearance to a glossy shine. > in the video?"
},
{
"from": "gpt",
"value": "<G>A person spea... | [
171.204,
178.144
] | [] | null |
6UQfcckT2oU | 331 | Autos & Vehicles | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/6UQfcckT2oU.mp4 | a2v | captioning | true | [
{
"from": "human",
"value": "<video>\nBased on the given audio cue < A person speaks, \"All right, I got the clear coat all on, and I'm just going to let it sit overnight, and then I'm going to sand it.\" >, identify the corresponding moment in the video and describe what is visually happening at that time.... | [
171.204,
178.144
] | [] | null |
fcfxiPZ3qS4 | 135 | Howto & Style | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/fcfxiPZ3qS4.mp4 | v2t | grounding | true | [
{
"from": "human",
"value": "<video>\nFind the video segment that corresponds to the given textual query < A person places and flattens cookie dough onto a baking sheet, followed by a display of baking instructions. Later, a woman in a kitchen presents a bowl of finished cookies. > and determine its start a... | [
85.6,
130.8
] | [] | null |
fcfxiPZ3qS4 | 135 | Howto & Style | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/fcfxiPZ3qS4.mp4 | a2t | grounding | true | [
{
"from": "human",
"value": "<video>\nOutput the start and end time of the video segment corresponding to this audio information: < Light, cheerful music plays throughout. An oven door closes. A woman speaks, \"That's it for this week. from lemon and chia seed cookies. The lemon gives the cookies a lovely f... | [
85.6,
130.8
] | [] | null |
fcfxiPZ3qS4 | 135 | Howto & Style | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/fcfxiPZ3qS4.mp4 | t2v | seg_captioning | true | [
{
"from": "human",
"value": "<video>\nOutput the visual information of the video segment from <G>."
},
{
"from": "gpt",
"value": "<G>An empty metal bowl on a marble counter is filled with spelt flour, baking soda, sea salt, coconut sugar, chia seeds, and lemon rind. These dry ingredients are the... | [
31.92,
85.56
] | [
31.92,
85.56
] | null |
fcfxiPZ3qS4 | 135 | Howto & Style | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/fcfxiPZ3qS4.mp4 | t2a | seg_captioning | true | [
{
"from": "human",
"value": "<video>\nWhat transpired from <G> in the video, according to the audio information?"
},
{
"from": "gpt",
"value": "<G>A female speaker says, \"First, add your flour,\" accompanied by the sound of powder being poured. She continues, \"your bicarb of soda, your sea sal... | [
31.92,
85.56
] | [
31.92,
85.56
] | null |
fcfxiPZ3qS4 | 135 | Howto & Style | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/fcfxiPZ3qS4.mp4 | dvc | captioning | true | [
{
"from": "human",
"value": "<video>\nCould you outline the incidents that happened during different time periods in the video?"
},
{
"from": "gpt",
"value": "<G><G><G><G><G><shift><G><G><G><G><G><shift>Visual: A stack of light brown, round cookies speckled with seeds, accompanied by whole yello... | [
0,
5,
5.04,
31.88,
31.92,
85.56,
85.6,
130.8,
130.84,
134.84,
0,
5,
5.04,
31.88,
31.92,
85.56,
85.6,
130.8,
130.84,
134.84
] | [] | [
"Visual: A stack of light brown, round cookies speckled with seeds, accompanied by whole yellow lemons. In the blurry background, a bowl of fruit is visible. Audio: Gentle, instrumental music plays.",
"Visual: A woman in a white knitted sweater stands in a kitchen with gray cabinets, a black countertop, and a sta... |
fcfxiPZ3qS4 | 135 | Howto & Style | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/fcfxiPZ3qS4.mp4 | v2a | captioning | true | [
{
"from": "human",
"value": "<video>\nIdentify the video segment where < A stack of light brown, round cookies speckled with seeds, accompanied by whole yellow lemons. In the blurry background, a bowl of fruit is visible. > occurs visually and infer the audio information present during that moment."
},
... | [
0,
5
] | [] | null |
fcfxiPZ3qS4 | 135 | Howto & Style | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/fcfxiPZ3qS4.mp4 | a2v | captioning | true | [
{
"from": "human",
"value": "<video>\nLocate the video segment associated with the audio event < Gentle, instrumental music plays. > and explain the visual details from an observerโs perspective."
},
{
"from": "gpt",
"value": "<G>A stack of light brown, round cookies speckled with seeds, accompa... | [
0,
5
] | [] | null |
o-XoUZSkKOU | 233 | Science & Technology | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/o-XoUZSkKOU.mp4 | v2t | grounding | true | [
{
"from": "human",
"value": "<video>\nDetect and report the start and end timestamps of the video segment that semantically matches the given textual query < A tiltrotor aircraft with navigation lights flies against a dark night sky. >."
},
{
"from": "gpt",
"value": "<G>"
}
] | [
170.666,
174.533
] | [] | null |
o-XoUZSkKOU | 233 | Science & Technology | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/o-XoUZSkKOU.mp4 | a2t | grounding | true | [
{
"from": "human",
"value": "<video>\nLocalize the audio content described by the given textual query < Electronic video game music is playing. > in the video, and output the start and end timestamps in seconds."
},
{
"from": "gpt",
"value": "<G>"
}
] | [
170.666,
174.533
] | [] | null |
o-XoUZSkKOU | 233 | Science & Technology | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/o-XoUZSkKOU.mp4 | t2v | seg_captioning | true | [
{
"from": "human",
"value": "<video>\nFocus on the video event in <G> and describe what is happening from the visual perspective."
},
{
"from": "gpt",
"value": "<G>A model V-22 Osprey aircraft with illuminated lights. A promotional card for \"Tiny Brite Lights.\""
}
] | [
0,
10.633
] | [
0,
10.633
] | null |
o-XoUZSkKOU | 233 | Science & Technology | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/o-XoUZSkKOU.mp4 | t2a | seg_captioning | true | [
{
"from": "human",
"value": "<video>\nWhat transpired from <G> in the video, according to the audio information?"
},
{
"from": "gpt",
"value": "<G>Upbeat, electronic music with a driving beat and a synthesized melody plays."
}
] | [
0,
10.633
] | [
0,
10.633
] | null |
o-XoUZSkKOU | 233 | Science & Technology | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/o-XoUZSkKOU.mp4 | dvc | captioning | true | [
{
"from": "human",
"value": "<video>\nCan you go through the video and describe what took place at different time intervals?"
},
{
"from": "gpt",
"value": "<G><G><G><G><G><G><G><G><G><G><G><G><G><G><shift><G><G><G><G><G><G><G><G><G><G><G><G><G><G><shift>Visual: A model V-22 Osprey aircraft with ... | [
0,
10.633,
10.666,
29.066,
29.1,
42.633,
42.666,
47.966,
48,
53.3,
53.333,
58.633,
58.666,
63.966,
64,
165.3,
165.333,
166.4,
166.433,
170.633,
170.666,
174.533,
174.566,
177.7,
177.733,
181.3,
181.333,
231.933,
0,
10.633,
10.666,
29.066,
29.1,
42.63... | [] | [
"Visual: A model V-22 Osprey aircraft with illuminated lights. A promotional card for \"Tiny Brite Lights.\" Audio: Upbeat, electronic music with a driving beat and a synthesized melody plays.",
"Visual: A close-up view of the internal components of a small drone or RC aircraft, featuring circuit boards, wires, a... |
o-XoUZSkKOU | 233 | Science & Technology | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/o-XoUZSkKOU.mp4 | v2a | captioning | true | [
{
"from": "human",
"value": "<video>\nIdentify the video segment where < A small aircraft, illuminated by red and green lights, takes off and ascends into the night sky, with some ground lights visible in the distance. > occurs visually and infer the audio information present during that moment."
},
{
... | [
166.433,
170.633
] | [] | null |
o-XoUZSkKOU | 233 | Science & Technology | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/o-XoUZSkKOU.mp4 | a2v | captioning | true | [
{
"from": "human",
"value": "<video>\nOutput the visual information of the video segment corresponding to this audio information: < Upbeat, electronic, video game-style music plays. >."
},
{
"from": "gpt",
"value": "<G>A small aircraft, illuminated by red and green lights, takes off and ascends ... | [
166.433,
170.633
] | [] | null |
4CywohtUCOo | 202 | News & Politics | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/4CywohtUCOo.mp4 | v2t | grounding | true | [
{
"from": "human",
"value": "<video>\nBetween which frames can we find: < A woman with a shopping cart stands in a supermarket aisle, examining products on refrigerated shelves. >?"
},
{
"from": "gpt",
"value": "<G>"
}
] | [
49.082,
51.051
] | [] | null |
4CywohtUCOo | 202 | News & Politics | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/4CywohtUCOo.mp4 | a2t | grounding | true | [
{
"from": "human",
"value": "<video>\nBetween which frames can we find: < A character speaking in a crisp voice, \"specifically at the grocery store.\" >?"
},
{
"from": "gpt",
"value": "<G>"
}
] | [
49.082,
51.051
] | [] | null |
4CywohtUCOo | 202 | News & Politics | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/4CywohtUCOo.mp4 | t2v | seg_captioning | true | [
{
"from": "human",
"value": "<video>\nProvide details about the visual events from <G> in the video."
},
{
"from": "gpt",
"value": "<G>Scenes alternate between a supermarket dairy aisle and a male presenter speaking in a studio setting. Graphics displaying food price inflation statistics are als... | [
70.403,
123.189
] | [
70.403,
123.189
] | null |
4CywohtUCOo | 202 | News & Politics | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/4CywohtUCOo.mp4 | t2a | seg_captioning | true | [
{
"from": "human",
"value": "<video>\nTell me about the audio events from <G> in the video."
},
{
"from": "gpt",
"value": "<G>A man speaks, \"...will be middle of the pack when it comes to inflation when you look at the historical averages. When you look overseas, the contrast becomes even more ... | [
70.403,
123.189
] | [
70.403,
123.189
] | null |
4CywohtUCOo | 202 | News & Politics | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/4CywohtUCOo.mp4 | dvc | captioning | true | [
{
"from": "human",
"value": "<video>\nCould you please detail the events that took place during different time segments in the video? List the events in the format: From xx to xx, event1. From xx to xx, event2. ..."
},
{
"from": "gpt",
"value": "<G><G><G><G><G><G><G><shift><G><G><G><G><G><G><G><... | [
0,
49.049,
49.082,
51.051,
51.084,
66.332,
66.366,
70.37,
70.403,
123.189,
123.223,
171.371,
171.404,
201.201,
0,
49.049,
49.082,
51.051,
51.084,
66.332,
66.366,
70.37,
70.403,
123.189,
123.223,
171.371,
171.404,
201.201
] | [] | [
"Visual: Studio presentations by two different anchors, each with a television monitor in the background displaying the \"Bottom Line\" title and sponsor logos. Animated title sequences featuring the \"Bottom Line\" logo and financial data also appear. A grocery store checkout scene is briefly shown. Audio: A man s... |
4CywohtUCOo | 202 | News & Politics | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/4CywohtUCOo.mp4 | v2a | captioning | true | [
{
"from": "human",
"value": "<video>\nOutput the audio information of the video segment corresponding to this visual information: < A male presenter in a studio introduces a program, followed by an animated title sequence displaying various agricultural scenes like fields of crops, tractors, and combine har... | [
171.404,
201.201
] | [] | null |
4CywohtUCOo | 202 | News & Politics | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/4CywohtUCOo.mp4 | a2v | captioning | true | [
{
"from": "human",
"value": "<video>\nFind the video segment that corresponds to the given textual query < A male voice says, \"Louisiana Agriculture, I'm Michael Danna. Thanks for watching, everybody. Hope to see you again right here next week.\" This is accompanied by upbeat, instrumental music featuring ... | [
171.404,
201.201
] | [] | null |
8lWgcQlmUUA | 304 | Education | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/8lWgcQlmUUA.mp4 | v2t | grounding | true | [
{
"from": "human",
"value": "<video>\nLocalize the visual content described by the given textual query < Animated text defines the word \"PERCEIVE\" with synonyms and meaning, followed by two examples demonstrating its use, accompanied by images of a man speaking at a podium and another man at a hearing. Th... | [
128.733,
176.733
] | [] | null |
8lWgcQlmUUA | 304 | Education | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/8lWgcQlmUUA.mp4 | a2t | grounding | true | [
{
"from": "human",
"value": "<video>\nFind the video segment that corresponds to the given textual query < A male speaker says, \"Perceive. The word perceive is a verb. The direct synonyms of perceive are understand, learn. In general context, the word perceive means come to realize or understand. We will s... | [
128.733,
176.733
] | [] | null |
8lWgcQlmUUA | 304 | Education | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/8lWgcQlmUUA.mp4 | t2v | seg_captioning | true | [
{
"from": "human",
"value": "<video>\nWhat was going on from <G> in the video? Reply should focus on visual information."
},
{
"from": "gpt",
"value": "<G>A white background with geometric patterns displays the word \"AFFIDAVIT\" with its definition and example sentences. This is followed by spl... | [
77.533,
128.7
] | [
77.533,
128.7
] | null |
8lWgcQlmUUA | 304 | Education | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/8lWgcQlmUUA.mp4 | t2a | seg_captioning | true | [
{
"from": "human",
"value": "<video>\nWhat transpired from <G> in the video, according to the audio information?"
},
{
"from": "gpt",
"value": "<G>A male voice speaks, \"Affidavit. The word affidavit is a noun. The direct synonyms of affidavit are legal document, testimony. In general context, t... | [
77.533,
128.7
] | [
77.533,
128.7
] | null |
8lWgcQlmUUA | 304 | Education | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/8lWgcQlmUUA.mp4 | dvc | captioning | true | [
{
"from": "human",
"value": "<video>\nCould you please detail the events that took place during different time segments in the video? List the events in the format: From xx to xx, event1. From xx to xx, event2. ..."
},
{
"from": "gpt",
"value": "<G><G><G><G><G><G><G><shift><G><G><G><G><G><G><G><... | [
0,
34.8,
34.833,
63.7,
63.733,
77.5,
77.533,
128.7,
128.733,
176.733,
176.766,
229.033,
229.066,
303.566,
0,
34.8,
34.833,
63.7,
63.733,
77.5,
77.533,
128.7,
128.733,
176.733,
176.766,
229.033,
229.066,
303.566
] | [] | [
"Visual: An animated logo reveal for \"EXAMBIN\" is shown on a white patterned background, transitioning to a title card displaying \"DAILY VIDEO VOCABULARY FROM EDITORIALS\" alongside \"THE HINDU\" logo on a red panel. This is followed by a red screen featuring blurred newspaper text and a date/topic slide, then a... |
8lWgcQlmUUA | 304 | Education | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/8lWgcQlmUUA.mp4 | v2a | captioning | true | [
{
"from": "human",
"value": "<video>\nWhat sounds or audio events accompany the visual scene described as < An animated logo reveal for \"EXAMBIN\" is shown on a white patterned background, transitioning to a title card displaying \"DAILY VIDEO VOCABULARY FROM EDITORIALS\" alongside \"THE HINDU\" logo on a ... | [
0,
34.8
] | [] | null |
8lWgcQlmUUA | 304 | Education | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/8lWgcQlmUUA.mp4 | a2v | captioning | true | [
{
"from": "human",
"value": "<video>\nFind the video segment that corresponds to the given textual query < A whooshing sound effect is followed by a male character speaking in a clear voice, \"Hello friends, welcome to daily video vocabulary. A daily dose of five to seven words from editorials of leading pu... | [
0,
34.8
] | [] | null |
4nrl7mwJRiA | 79 | People & Blogs | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/4nrl7mwJRiA.mp4 | v2t | grounding | true | [
{
"from": "human",
"value": "<video>\nFind the video segment that corresponds to the given textual query < A stew-like dish with noodles, shredded meat, and diced vegetables is stirred and served from a blue crock-pot. > and determine its start and end seconds."
},
{
"from": "gpt",
"value": "<G>... | [
52.685,
62.996
] | [] | null |
4nrl7mwJRiA | 79 | People & Blogs | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/4nrl7mwJRiA.mp4 | a2t | grounding | true | [
{
"from": "human",
"value": "<video>\nFind the video segment that corresponds to the given textual query < A female speaker with a clear voice says, \"So once your noodles are done, you're going to want to add them into the crock pot and mix it up with your chicken and your vegetables. Then you're going to ... | [
52.685,
62.996
] | [] | null |
4nrl7mwJRiA | 79 | People & Blogs | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/4nrl7mwJRiA.mp4 | t2v | seg_captioning | true | [
{
"from": "human",
"value": "<video>\nWhat was going on from <G> in the video? Reply should focus on visual information."
},
{
"from": "gpt",
"value": "<G>A person holds up a carton of chicken broth, then pours from a measuring cup into a pot."
}
] | [
10.143,
12.312
] | [
10.143,
12.312
] | null |
4nrl7mwJRiA | 79 | People & Blogs | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/4nrl7mwJRiA.mp4 | t2a | seg_captioning | true | [
{
"from": "human",
"value": "<video>\nTell me about the audio events from <G> in the video."
},
{
"from": "gpt",
"value": "<G>A person speaks, saying, \"garlic, chicken broth, and\"."
}
] | [
10.143,
12.312
] | [
10.143,
12.312
] | null |
4nrl7mwJRiA | 79 | People & Blogs | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/4nrl7mwJRiA.mp4 | dvc | captioning | true | [
{
"from": "human",
"value": "<video>\nCould you please detail the events that took place during different time segments in the video? List the events in the format: From xx to xx, event1. From xx to xx, event2. ..."
},
{
"from": "gpt",
"value": "<G><G><G><G><G><G><G><G><G><G><G><G><G><shift><G><... | [
0,
3.97,
4.004,
6.206,
6.239,
10.11,
10.143,
12.312,
12.345,
14.848,
14.881,
29.162,
29.195,
32.432,
32.465,
52.652,
52.685,
62.996,
63.029,
65.365,
65.398,
68.334,
68.368,
74.14,
74.174,
78.812,
0,
3.97,
4.004,
6.206,
6.239,
10.11,
10.143,
12.312,
... | [] | [
"Visual: White text on a black background displaying \"Chicken noodle soup! cook on low 6-8 hours\". Audio: A man says, \"So she and she just\".",
"Visual: A person opens a package of food and pours its contents into a slow cooker in an indoor setting. Audio: A speaker says, \"So what you're going to want to add.... |
4nrl7mwJRiA | 79 | People & Blogs | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/4nrl7mwJRiA.mp4 | v2a | captioning | true | [
{
"from": "human",
"value": "<video>\nWhat sounds or audio events accompany the visual scene described as < A person adds a pat of butter to a slow cooker filled with a stew-like dish and then stirs it with a slotted spoon. > in the video?"
},
{
"from": "gpt",
"value": "<G>A person speaks, \"and... | [
63.029,
65.365
] | [] | null |
4nrl7mwJRiA | 79 | People & Blogs | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/4nrl7mwJRiA.mp4 | a2v | captioning | true | [
{
"from": "human",
"value": "<video>\nBased on the given audio cue < A person speaks, \"and add in, you don't want to add a lot.\" >, identify the corresponding moment in the video and describe what is visually happening at that time."
},
{
"from": "gpt",
"value": "<G>A person adds a pat of butt... | [
63.029,
65.365
] | [] | null |
mc1B7snTM2U | 72 | News & Politics | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/mc1B7snTM2U.mp4 | v2t | grounding | true | [
{
"from": "human",
"value": "<video>\nLocalize the visual content described by the given textual query < A large assembly of people seated in a spacious hall with prominent news channel graphics overlaid at the bottom. > in the video, and output the start and end timestamps in seconds."
},
{
"from":... | [
19.5,
22.466
] | [] | null |
mc1B7snTM2U | 72 | News & Politics | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/mc1B7snTM2U.mp4 | a2t | grounding | true | [
{
"from": "human",
"value": "<video>\nBetween which frames can we find: < A female voice speaks, \"During his campaign, Trump had dis...\" >?"
},
{
"from": "gpt",
"value": "<G>"
}
] | [
19.5,
22.466
] | [] | null |
mc1B7snTM2U | 72 | News & Politics | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/mc1B7snTM2U.mp4 | t2v | seg_captioning | true | [
{
"from": "human",
"value": "<video>\nFocus on the video event in <G> and describe what is happening from the visual perspective."
},
{
"from": "gpt",
"value": "<G>An industrial complex or refinery with numerous large towers, interconnected pipes, and storage tanks under a hazy sky. Some buildin... | [
31.5,
37.533
] | [
31.5,
37.533
] | null |
mc1B7snTM2U | 72 | News & Politics | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/mc1B7snTM2U.mp4 | t2a | seg_captioning | true | [
{
"from": "human",
"value": "<video>\nFocus on the video event in <G> and describe what is happening from audio perspective."
},
{
"from": "gpt",
"value": "<G>A female voice says, \"...a program in return for lifted sanctions. Rouhani said he will attend a meeting of an experts committee.\""
}... | [
31.5,
37.533
] | [
31.5,
37.533
] | null |
mc1B7snTM2U | 72 | News & Politics | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/mc1B7snTM2U.mp4 | dvc | captioning | true | [
{
"from": "human",
"value": "<video>\nCould you please detail the events that took place during different time segments in the video? List the events in the format: From xx to xx, event1. From xx to xx, event2. ..."
},
{
"from": "gpt",
"value": "<G><G><G><G><G><G><G><G><G><G><G><G><G><shift><G><... | [
0,
7.966,
8,
10.733,
10.766,
13.733,
13.766,
16.733,
16.766,
19.466,
19.5,
22.466,
22.5,
25.466,
25.5,
28.466,
28.5,
31.466,
31.5,
37.533,
37.566,
40.166,
40.2,
42.333,
42.366,
71.133,
0,
7.966,
8,
10.733,
10.766,
13.733,
13.766,
16.733,
16.766,
... | [] | [
"Visual: A female news anchor in a blue jacket presents a broadcast. Behind her, a visual graphic displays a map of Iran with the word \"IRAN\" in large red letters. A brick wall labeled \"SANCTIONS\" and topped with barbed wire is positioned in front of the map. CCTV NEWS branding and a \"NEWS DESK\" logo are visi... |
mc1B7snTM2U | 72 | News & Politics | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/mc1B7snTM2U.mp4 | v2a | captioning | true | [
{
"from": "human",
"value": "<video>\nWhat sounds or audio events accompany the visual scene described as < A large audience, mostly men, is seated in an auditorium or conference hall, with news graphics overlaid at the bottom. > in the video?"
},
{
"from": "gpt",
"value": "<G>A person speaks, \... | [
13.766,
16.733
] | [] | null |
mc1B7snTM2U | 72 | News & Politics | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/mc1B7snTM2U.mp4 | a2v | captioning | true | [
{
"from": "human",
"value": "<video>\nLocate the video segment associated with the audio event < A person speaks, \"decision last month to extend the Iran sanctions\". > and explain the visual details from an observerโs perspective."
},
{
"from": "gpt",
"value": "<G>A large audience, mostly men,... | [
13.766,
16.733
] | [] | null |
RkPBozKned0 | 97 | News & Politics | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/RkPBozKned0.mp4 | v2t | grounding | true | [
{
"from": "human",
"value": "<video>\nOutput the start and end time of the video segment corresponding to this video information: < A wet road with green grassy shoulders under an overcast sky, overlaid with a news banner reporting \"SEARCHING FOR HIT AND RUN DRIVER\" and \"WOMAN KILLED WHILE WALKING IN LOX... | [
52.819,
59.325
] | [] | null |
RkPBozKned0 | 97 | News & Politics | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/RkPBozKned0.mp4 | a2t | grounding | true | [
{
"from": "human",
"value": "<video>\nGive you a textual query: < A man speaks, \"It's actually scary to hear about what happened because he and many other families actually walk in that same exact spot all the time,\" as a vehicle drives by. > When does the described content occur in the video? Please retu... | [
52.819,
59.325
] | [] | null |
RkPBozKned0 | 97 | News & Politics | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/RkPBozKned0.mp4 | t2v | seg_captioning | true | [
{
"from": "human",
"value": "<video>\nFocus on the video event in <G> and describe what is happening from the visual perspective."
},
{
"from": "gpt",
"value": "<G>A blurry white car drives on a wet road in the rain, followed by an older man with glasses speaking from inside a car. A news ticker... | [
59.359,
71.438
] | [
59.359,
71.438
] | null |
RkPBozKned0 | 97 | News & Politics | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/RkPBozKned0.mp4 | t2a | seg_captioning | true | [
{
"from": "human",
"value": "<video>\nFocus on the video event in <G> and describe what is happening from audio perspective."
},
{
"from": "gpt",
"value": "<G>A whooshing sound is followed by a man speaking in a clear voice, \"Of course, I have to think about all the families that live in here. ... | [
59.359,
71.438
] | [
59.359,
71.438
] | null |
RkPBozKned0 | 97 | News & Politics | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/RkPBozKned0.mp4 | dvc | captioning | true | [
{
"from": "human",
"value": "<video>\nCould you please detail the events that took place during different time segments in the video? List the events in the format: From xx to xx, event1. From xx to xx, event2. ..."
},
{
"from": "gpt",
"value": "<G><G><G><G><G><shift><G><G><G><G><G><shift>Visual... | [
0,
44.778,
44.811,
52.786,
52.819,
59.325,
59.359,
71.438,
71.471,
96.83,
0,
44.778,
44.811,
52.786,
52.819,
59.325,
59.359,
71.438,
71.471,
96.83
] | [] | [
"Visual: A news report features a female anchor in a studio, followed by a male reporter outdoors on the side of a road. On-screen graphics display headlines about a hit-and-run incident and a woman killed while walking. Photos of a smiling woman, identified as the victim, are shown. Audio: A woman speaks, \"The se... |
RkPBozKned0 | 97 | News & Politics | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/RkPBozKned0.mp4 | v2a | captioning | true | [
{
"from": "human",
"value": "<video>\nOutput the audio information of the video segment corresponding to this visual information: < A news report features a female anchor in a studio, followed by a male reporter outdoors on the side of a road. On-screen graphics display headlines about a hit-and-run inciden... | [
0,
44.778
] | [] | null |
RkPBozKned0 | 97 | News & Politics | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/RkPBozKned0.mp4 | a2v | captioning | true | [
{
"from": "human",
"value": "<video>\nLocate the video segment associated with the audio event < A woman speaks, \"The search is on this midday for the driver investigators say hit and killed a woman who was walking on the side of the road and then drove away. News Channel 5's Jeff Scripeck joins us live at... | [
0,
44.778
] | [] | null |
ri9oBA7FVfE | 324 | Sports | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/ri9oBA7FVfE.mp4 | v2t | grounding | true | [
{
"from": "human",
"value": "<video>\nFind the video segment that corresponds to the given textual query < Cave divers exploring an underwater cave system, then trekking through a dense forest with their dive equipment. > and determine its start and end seconds."
},
{
"from": "gpt",
"value": "<G... | [
279.579,
298.598
] | [] | null |
ri9oBA7FVfE | 324 | Sports | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/ri9oBA7FVfE.mp4 | a2t | grounding | true | [
{
"from": "human",
"value": "<video>\nLocalize the audio content described by the given textual query < Rhythmic electronic music with a steady beat plays. > in the video, and output the start and end timestamps in seconds."
},
{
"from": "gpt",
"value": "<G>"
}
] | [
279.579,
298.598
] | [] | null |
ri9oBA7FVfE | 324 | Sports | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/ri9oBA7FVfE.mp4 | t2v | seg_captioning | true | [
{
"from": "human",
"value": "<video>\nProvide details about the visual events from <G> in the video."
},
{
"from": "gpt",
"value": "<G>A video featuring information about a GoPro Hero 3+ camera setup for diving, including its housing and settings, alongside logos and details of various diving or... | [
298.631,
323.189
] | [
298.631,
323.189
] | null |
ri9oBA7FVfE | 324 | Sports | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/ri9oBA7FVfE.mp4 | t2a | seg_captioning | true | [
{
"from": "human",
"value": "<video>\nFocus on the video event in <G> and describe what is happening from audio perspective."
},
{
"from": "gpt",
"value": "<G>Atmospheric electronic music plays, featuring a pulsing beat and layers of synth melodies."
}
] | [
298.631,
323.189
] | [
298.631,
323.189
] | null |
ri9oBA7FVfE | 324 | Sports | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/ri9oBA7FVfE.mp4 | dvc | captioning | true | [
{
"from": "human",
"value": "<video>\nI'd like to know what events transpired during specific time intervals in the video. Could you please elaborate?"
},
{
"from": "gpt",
"value": "<G><G><G><G><G><G><G><G><shift><G><G><G><G><G><G><G><G><shift>Visual: Screens displaying logos and text related to... | [
0,
18.151,
18.184,
127.861,
127.894,
191.057,
191.09,
200.8,
200.833,
222.455,
222.488,
279.545,
279.579,
298.598,
298.631,
323.189,
0,
18.151,
18.184,
127.861,
127.894,
191.057,
191.09,
200.8,
200.833,
222.455,
222.488,
279.545,
279.579,
298.598,
298.631,... | [] | [
"Visual: Screens displaying logos and text related to cave diving businesses, photographers, and video equipment specifications. Audio: Electronic music with a rhythmic beat and a synthesized melody is playing.",
"Visual: People preparing for a dive, then entering and exploring an underwater cave system filled wi... |
ri9oBA7FVfE | 324 | Sports | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/ri9oBA7FVfE.mp4 | v2a | captioning | true | [
{
"from": "human",
"value": "<video>\nIdentify the video segment where < Cave divers explore an intricate underwater cavern filled with clear, luminous green and blue water, navigating among dramatic stalactite and stalagmite formations. > occurs visually and infer the audio information present during that ... | [
222.488,
279.545
] | [] | null |
ri9oBA7FVfE | 324 | Sports | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/ri9oBA7FVfE.mp4 | a2v | captioning | true | [
{
"from": "human",
"value": "<video>\nWhen the audio indicates < A melancholic piano melody plays, accompanied by synth pads and a deep, booming drum sound. A rhythmic electronic beat begins, with synth arpeggios and various electronic textures building throughout the track. >, what visual scene or actions ... | [
222.488,
279.545
] | [] | null |
tr7uJ-IxjLg | 81 | Music | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/tr7uJ-IxjLg.mp4 | v2t | grounding | true | [
{
"from": "human",
"value": "<video>\nDetect and report the start and end timestamps of the video segment that semantically matches the given textual query < A pedestrian traffic light cycling through stop and walk signals in various colors (red, green, blue) with a numerical countdown, set against a cloudy... | [
41.074,
44.077
] | [] | null |
tr7uJ-IxjLg | 81 | Music | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/tr7uJ-IxjLg.mp4 | a2t | grounding | true | [
{
"from": "human",
"value": "<video>\nGive you a textual query: < A series of electronic beeps followed by a sustained electronic tone. > When does the described content occur in the video? Please return the timestamp in seconds."
},
{
"from": "gpt",
"value": "<G>"
}
] | [
41.074,
44.077
] | [] | null |
tr7uJ-IxjLg | 81 | Music | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/tr7uJ-IxjLg.mp4 | t2v | seg_captioning | true | [
{
"from": "human",
"value": "<video>\nWhat transpired from <G> in the video, according to the visual information?"
},
{
"from": "gpt",
"value": "<G>A pedestrian traffic light with a countdown timer cycles through red, green, and blue signals, accompanied by text overlays \"Round 2\" and \"ๆฅๆฌ่ชๅญๅน\... | [
30.03,
33.033
] | [
30.03,
33.033
] | null |
tr7uJ-IxjLg | 81 | Music | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/tr7uJ-IxjLg.mp4 | t2a | seg_captioning | true | [
{
"from": "human",
"value": "<video>\nWhat transpired from <G> in the video, according to the audio information?"
},
{
"from": "gpt",
"value": "<G>A series of beeping sounds is followed by a sustained musical chord."
}
] | [
30.03,
33.033
] | [
30.03,
33.033
] | null |
tr7uJ-IxjLg | 81 | Music | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/tr7uJ-IxjLg.mp4 | dvc | captioning | true | [
{
"from": "human",
"value": "<video>\nCan you go through the video and describe what took place at different time intervals?"
},
{
"from": "gpt",
"value": "<G><G><G><G><G><G><G><G><G><G><shift><G><G><G><G><G><G><G><G><G><G><shift>Visual: A person with long dark hair and glasses wears headphones ... | [
0,
12.979,
13.013,
18.918,
18.952,
21.988,
22.022,
29.996,
30.03,
33.033,
33.066,
41.041,
41.074,
44.077,
44.11,
57.09,
57.123,
67.1,
67.133,
80.046,
0,
12.979,
13.013,
18.918,
18.952,
21.988,
22.022,
29.996,
30.03,
33.033,
33.066,
41.041,
41.074,
44... | [] | [
"Visual: A person with long dark hair and glasses wears headphones while looking at a device, with red text overlaid on the scene. Audio: Chimes ring as a female voice says in Japanese, \"4ๆ้็ฎ ่ฑไผ่ฉฑ.\" A second female voice then says, \"Three rounds listening practice.\" The chimes continue to play.",
"Visual: A st... |
tr7uJ-IxjLg | 81 | Music | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/tr7uJ-IxjLg.mp4 | v2a | captioning | true | [
{
"from": "human",
"value": "<video>\nOutput the audio information of the video segment corresponding to this visual information: < A person dressed in a Grim Reaper costume with a skull mask and a fake axe in their head stands in a doorway decorated with string lights, while various English and Japanese te... | [
44.11,
57.09
] | [] | null |
tr7uJ-IxjLg | 81 | Music | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/tr7uJ-IxjLg.mp4 | a2v | captioning | true | [
{
"from": "human",
"value": "<video>\nFind the video segment that corresponds to the given textual query < A character speaks, \"Hey, don't yell at Mr. Q for making some mistakes. You've got to cut him some slack. There's no perfect person out there.\" > and tell me what was going from visual perspective."
... | [
44.11,
57.09
] | [] | null |
Kjr32fF2kSQ | 164 | Film & Animation | /storage/wenzheng/dataset/omni/ChronusAV/video/train_resized/Kjr32fF2kSQ.mp4 | v2t | grounding | true | [
{
"from": "human",
"value": "<video>\nFind the video segment that corresponds to the given textual query < A person heavily dressed in red and covered in snow lies on a treacherous, icy mountain slope, with vast snowy peaks and clouds in the background. A woman with dark hair speaks on a cordless phone indo... | [
140.849,
147.397
] | [] | null |
End of preview.
๐ฌ PadCaptioner
Parallelized Autoregressive Decoding for Omni-Modal Dense Video Captioning
Wenzheng Zeng, Siyi Jiao, Chen Gao, Hwee Tou Ng, Mike Zheng Shou
National University of Singapore
ECCV 2026
Here we provide the data (annotation files) used in PadCaptioner.
๐ Citation
If you find our work useful, please kindly cite:
@inproceedings{padcaptioner,
title={Parallelized Autoregressive Decoding for Omni-Modal Dense Video Captioning},
author={Zeng, Wenzheng and Jiao, Siyi and Gao, Chen and Ng, Hwee Tou and Shou, Mike Zheng},
booktitle={European Conference on Computer Vision (ECCV)},
year={2026}
}
- Downloads last month
- 25