task
stringclasses 1
value | source
stringclasses 1
value | video
stringlengths 37
38
| duration
int64 120
150
| tgt
listlengths 1
24
| conversations
listlengths 2
2
|
|---|---|---|---|---|---|
tvg
|
qvhighlights
|
qvh/videos/u6qkkpoDwvI_60.0_210.0.mp4
| 150
|
[
[
0,
80
]
] |
[
{
"from": "human",
"value": "<image>\nFind the video segment that corresponds to the given textual query 'A man doing a food review' and determine its start and end seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/u6qkkpoDwvI_360.0_510.0.mp4
| 150
|
[
[
104,
150
]
] |
[
{
"from": "human",
"value": "<image>\nGiven a textual query: 'An Asian man sitting on his bed next to a window talking to the camera', when does the described content occur in the video? Please return the timestamp in seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/u6qkkpoDwvI_210.0_360.0.mp4
| 150
|
[
[
36,
52
]
] |
[
{
"from": "human",
"value": "<image>\nGiven a textual query: 'Man holds up a bowl of pasta', when does the described content occur in the video? Please return the timestamp in seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/BraQgRtlwjs_60.0_210.0.mp4
| 150
|
[
[
0,
150
]
] |
[
{
"from": "human",
"value": "<image>\nThe given natural language query 'A man talks to the camera holding a book and a microphone' is semantically aligned with a video moment, please give the start time and end time of the video moment.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/BraQgRtlwjs_210.0_360.0.mp4
| 150
|
[
[
22,
54
]
] |
[
{
"from": "human",
"value": "<image>\nDetect and report the start and end timestamps of the video segment that semantically matches the given textual query 'Man in white top plays guitar'.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/BraQgRtlwjs_360.0_510.0.mp4
| 150
|
[
[
0,
46
]
] |
[
{
"from": "human",
"value": "<image>\nDetect and report the start and end timestamps of the video segment that semantically matches the given textual query 'A young woman in a red costume sings a traditional folk song'.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/BraQgRtlwjs_660.0_810.0.mp4
| 150
|
[
[
0,
116
]
] |
[
{
"from": "human",
"value": "<image>\nDetect and report the start and end timestamps of the video segment that semantically matches the given textual query 'Man in white top talks into a mic'.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/BraQgRtlwjs_510.0_660.0.mp4
| 150
|
[
[
32,
78
]
] |
[
{
"from": "human",
"value": "<image>\nFind the video segment that corresponds to the given textual query 'A little girl dances and sings' and determine its start and end seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/0cDIDBCpWVA_210.0_360.0.mp4
| 150
|
[
[
0,
150
]
] |
[
{
"from": "human",
"value": "<image>\nDetect and report the start and end timestamps of the video segment that semantically matches the given textual query 'South Indian officials press conference about relief efforts in flooding area'.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/Tm54Ci2ut4M_360.0_510.0.mp4
| 150
|
[
[
112,
120
],
[
124,
150
]
] |
[
{
"from": "human",
"value": "<image>\nFind the video segment that corresponds to the given textual query 'man fixing her little baby in scroller' and determine its start and end seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/Tm54Ci2ut4M_210.0_360.0.mp4
| 150
|
[
[
0,
22
],
[
74,
78
],
[
86,
96
],
[
98,
134
],
[
136,
138
],
[
140,
148
]
] |
[
{
"from": "human",
"value": "<image>\nLocate the visual content mentioned in the text query 'Cute videos of a baby' within the video using timestamps.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/Tm54Ci2ut4M_60.0_210.0.mp4
| 150
|
[
[
118,
132
]
] |
[
{
"from": "human",
"value": "<image>\nLocalize the visual content described by the given textual query 'A family is checking on a baby laying in it's carrier on a booth at a restaurant' in the video, and output the start and end timestamps in seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/Tm54Ci2ut4M_510.0_660.0.mp4
| 150
|
[
[
46,
64
]
] |
[
{
"from": "human",
"value": "<image>\nThe given natural language query 'A baby is jumping while being secured by a colorful protective bumper' is semantically aligned with a video moment, please give the start time and end time of the video moment.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/Tm54Ci2ut4M_660.0_810.0.mp4
| 150
|
[
[
0,
150
]
] |
[
{
"from": "human",
"value": "<image>\nDetect and report the start and end timestamps of the video segment that semantically matches the given textual query 'A couple is playing Bean Boozled jelly bean game'.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/QmYWUV957bw_60.0_210.0.mp4
| 150
|
[
[
134,
150
]
] |
[
{
"from": "human",
"value": "<image>\nLocate the visual content mentioned in the text query 'Old man holds up a Hawaiian shirt' within the video using timestamps.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/QmYWUV957bw_360.0_510.0.mp4
| 150
|
[
[
42,
134
]
] |
[
{
"from": "human",
"value": "<image>\nLocate the visual content mentioned in the text query 'Father and son having great time at Florida beach' within the video using timestamps.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/QmYWUV957bw_210.0_360.0.mp4
| 150
|
[
[
24,
26
],
[
54,
114
]
] |
[
{
"from": "human",
"value": "<image>\nLocate the visual content mentioned in the text query 'Two men sitting on armchairs next to each other holding new shirts' within the video using timestamps.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/QmYWUV957bw_510.0_660.0.mp4
| 150
|
[
[
24,
70
]
] |
[
{
"from": "human",
"value": "<image>\nLocate the visual content mentioned in the text query 'A woman and her parents pose for photos on a beach with breaking waves at sunset' within the video using timestamps.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/A5i_pxq3PKI_210.0_360.0.mp4
| 150
|
[
[
72,
94
]
] |
[
{
"from": "human",
"value": "<image>\nLocate the visual content mentioned in the text query 'Photo of Austan Goolsbee next to a icon indicating his audio is playing' within the video using timestamps.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/A5i_pxq3PKI_360.0_510.0.mp4
| 150
|
[
[
2,
36
]
] |
[
{
"from": "human",
"value": "<image>\nDetect and report the start and end timestamps of the video segment that semantically matches the given textual query 'News caster is showing the insight view of the studio'.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/A5i_pxq3PKI_60.0_210.0.mp4
| 150
|
[
[
28,
48
]
] |
[
{
"from": "human",
"value": "<image>\nLocalize the visual content described by the given textual query 'Different types and styles of stained glass are being explained' in the video, and output the start and end timestamps in seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/NMksNT3ouG4_360.0_510.0.mp4
| 150
|
[
[
58,
60
],
[
70,
76
],
[
86,
90
],
[
100,
114
],
[
128,
150
]
] |
[
{
"from": "human",
"value": "<image>\nGiven a textual query: 'Man and woman are eating with chopsticks', when does the described content occur in the video? Please return the timestamp in seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/NMksNT3ouG4_210.0_360.0.mp4
| 150
|
[
[
80,
112
]
] |
[
{
"from": "human",
"value": "<image>\nThe given natural language query 'Girl filming boy friend playing piano' is semantically aligned with a video moment, please give the start time and end time of the video moment.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/NMksNT3ouG4_60.0_210.0.mp4
| 150
|
[
[
62,
106
],
[
108,
118
],
[
120,
128
]
] |
[
{
"from": "human",
"value": "<image>\nThe given natural language query 'Video editing software showing the process of making a video' is semantically aligned with a video moment, please give the start time and end time of the video moment.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/NMksNT3ouG4_510.0_660.0.mp4
| 150
|
[
[
120,
134
]
] |
[
{
"from": "human",
"value": "<image>\nGiven a textual query: 'Woman holds up a steady cam to the mirror', when does the described content occur in the video? Please return the timestamp in seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/zTF6iU4SLmM_360.0_510.0.mp4
| 150
|
[
[
122,
144
]
] |
[
{
"from": "human",
"value": "<image>\nDetect and report the start and end timestamps of the video segment that semantically matches the given textual query 'A young girl watching a show on an airplane TV'.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/zTF6iU4SLmM_210.0_360.0.mp4
| 150
|
[
[
0,
150
]
] |
[
{
"from": "human",
"value": "<image>\nThe given natural language query 'A girl unpacking her travel items' is semantically aligned with a video moment, please give the start time and end time of the video moment.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/zTF6iU4SLmM_60.0_210.0.mp4
| 150
|
[
[
98,
100
],
[
114,
150
]
] |
[
{
"from": "human",
"value": "<image>\nThe given natural language query 'Woman looks at the pockets of her backpack' is semantically aligned with a video moment, please give the start time and end time of the video moment.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/p8FS8kq5YIU_210.0_360.0.mp4
| 150
|
[
[
24,
48
]
] |
[
{
"from": "human",
"value": "<image>\nFind the video segment that corresponds to the given textual query 'A woman in a car picks a coffee mug and places it down' and determine its start and end seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/p8FS8kq5YIU_60.0_210.0.mp4
| 150
|
[
[
0,
116
],
[
120,
150
]
] |
[
{
"from": "human",
"value": "<image>\nThe given natural language query 'Girl giving reviews on vlog while talking to her friend' is semantically aligned with a video moment, please give the start time and end time of the video moment.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/p8FS8kq5YIU_360.0_510.0.mp4
| 150
|
[
[
0,
150
]
] |
[
{
"from": "human",
"value": "<image>\nLocalize the visual content described by the given textual query 'A young couple sit in a sunlit room and talk' in the video, and output the start and end timestamps in seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/p8FS8kq5YIU_660.0_810.0.mp4
| 150
|
[
[
36,
50
]
] |
[
{
"from": "human",
"value": "<image>\nDetect and report the start and end timestamps of the video segment that semantically matches the given textual query 'Vlogger installs a entire couch sectional by the pool'.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/p8FS8kq5YIU_510.0_660.0.mp4
| 150
|
[
[
0,
16
]
] |
[
{
"from": "human",
"value": "<image>\nFind the video segment that corresponds to the given textual query 'a man and woman address the camera together' and determine its start and end seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/5-pHuu3MHPU_360.0_510.0.mp4
| 150
|
[
[
78,
106
]
] |
[
{
"from": "human",
"value": "<image>\nLocate the visual content mentioned in the text query 'A kid inside a big balloon on water' within the video using timestamps.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/5-pHuu3MHPU_360.0_510.0.mp4
| 150
|
[
[
78,
108
]
] |
[
{
"from": "human",
"value": "<image>\nFind the video segment that corresponds to the given textual query 'A child getting in a green ball pool toy and playing in it at a pool' and determine its start and end seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/5-pHuu3MHPU_60.0_210.0.mp4
| 150
|
[
[
90,
112
],
[
118,
120
]
] |
[
{
"from": "human",
"value": "<image>\nLocate the visual content mentioned in the text query 'Girl in pink top holds parent's hand at the beach' within the video using timestamps.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/5-pHuu3MHPU_210.0_360.0.mp4
| 150
|
[
[
114,
144
]
] |
[
{
"from": "human",
"value": "<image>\nLocalize the visual content described by the given textual query 'A young girl in pink is eating at the table in the restaurant and giving her thoughts on the food' in the video, and output the start and end timestamps in seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/5-pHuu3MHPU_60.0_210.0.mp4
| 150
|
[
[
92,
112
]
] |
[
{
"from": "human",
"value": "<image>\nDetect and report the start and end timestamps of the video segment that semantically matches the given textual query 'Dad is holding his daughter's hand'.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/5-pHuu3MHPU_210.0_360.0.mp4
| 150
|
[
[
100,
114
]
] |
[
{
"from": "human",
"value": "<image>\nThe given natural language query 'Girl plays in a hot tub' is semantically aligned with a video moment, please give the start time and end time of the video moment.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/5-pHuu3MHPU_660.0_810.0.mp4
| 150
|
[
[
58,
96
]
] |
[
{
"from": "human",
"value": "<image>\nFind the video segment that corresponds to the given textual query 'Two children are parasailing at the back of a boat' and determine its start and end seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/5-pHuu3MHPU_510.0_660.0.mp4
| 150
|
[
[
0,
14
]
] |
[
{
"from": "human",
"value": "<image>\nLocate the visual content mentioned in the text query 'A little girl with navy dress is eating her lunch on the table' within the video using timestamps.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/5-pHuu3MHPU_660.0_810.0.mp4
| 150
|
[
[
58,
96
]
] |
[
{
"from": "human",
"value": "<image>\nGiven a textual query: 'A little girl tries parasailing on a windy overcast day', when does the described content occur in the video? Please return the timestamp in seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/ghFXlxuca_U_60.0_210.0.mp4
| 150
|
[
[
28,
30
],
[
42,
46
],
[
50,
52
],
[
54,
58
],
[
62,
70
],
[
76,
80
],
[
100,
106
],
[
112,
118
]
] |
[
{
"from": "human",
"value": "<image>\nLocalize the visual content described by the given textual query 'a woman with black shirt and wearing a toque is talking in front of camera about what they brought during her trip such ass shoes, swimwear in iceland' in the video, and output the start and end timestamps in seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/ghFXlxuca_U_210.0_360.0.mp4
| 150
|
[
[
0,
30
],
[
32,
42
]
] |
[
{
"from": "human",
"value": "<image>\nFind the video segment that corresponds to the given textual query 'Woman in beanie takes off her backpack and lays it across her lap' and determine its start and end seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/ghFXlxuca_U_360.0_510.0.mp4
| 150
|
[
[
22,
32
],
[
44,
56
],
[
60,
70
]
] |
[
{
"from": "human",
"value": "<image>\nLocalize the visual content described by the given textual query 'A young woman sitting on a bed grabbing some books and showing them to the camera' in the video, and output the start and end timestamps in seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/KXAbVkEWsPk_60.0_210.0.mp4
| 150
|
[
[
22,
24
],
[
26,
92
],
[
94,
100
],
[
104,
114
],
[
116,
128
],
[
130,
150
]
] |
[
{
"from": "human",
"value": "<image>\nLocate the visual content mentioned in the text query 'Father and son having some lockdown conversation during drive' within the video using timestamps.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/KXAbVkEWsPk_360.0_510.0.mp4
| 150
|
[
[
0,
2
],
[
4,
8
],
[
14,
24
],
[
58,
80
],
[
82,
104
],
[
106,
108
],
[
112,
120
],
[
146,
150
]
] |
[
{
"from": "human",
"value": "<image>\nFind the video segment that corresponds to the given textual query 'Beaded man is talking to the camera while driving' and determine its start and end seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/KXAbVkEWsPk_210.0_360.0.mp4
| 150
|
[
[
0,
10
],
[
32,
44
],
[
48,
56
]
] |
[
{
"from": "human",
"value": "<image>\nGiven a textual query: 'A video is showing many buildings, houses, streets and so on', when does the described content occur in the video? Please return the timestamp in seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/KXAbVkEWsPk_510.0_660.0.mp4
| 150
|
[
[
54,
150
]
] |
[
{
"from": "human",
"value": "<image>\nLocalize the visual content described by the given textual query 'Man wearing a Friends shirt talk to his friend wearing a blue shirt' in the video, and output the start and end timestamps in seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/K81I34r_o6w_210.0_360.0.mp4
| 150
|
[
[
0,
14
]
] |
[
{
"from": "human",
"value": "<image>\nGiven a textual query: 'Woman is riding in the back of a van', when does the described content occur in the video? Please return the timestamp in seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/K81I34r_o6w_60.0_210.0.mp4
| 150
|
[
[
50,
72
]
] |
[
{
"from": "human",
"value": "<image>\nLocate the visual content mentioned in the text query 'Women have a conversation lounging in their hotel room' within the video using timestamps.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/K81I34r_o6w_360.0_510.0.mp4
| 150
|
[
[
136,
150
]
] |
[
{
"from": "human",
"value": "<image>\nGiven a textual query: 'A woman is petting a camel that is laying in the sand', when does the described content occur in the video? Please return the timestamp in seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/ev1s6q2NopA_60.0_210.0.mp4
| 150
|
[
[
94,
122
],
[
124,
150
]
] |
[
{
"from": "human",
"value": "<image>\nFind the video segment that corresponds to the given textual query 'A guy talking from a top floor of a house' and determine its start and end seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/ev1s6q2NopA_210.0_360.0.mp4
| 150
|
[
[
0,
86
],
[
112,
136
]
] |
[
{
"from": "human",
"value": "<image>\nFind the video segment that corresponds to the given textual query 'Man in plaid shirt gives a presentation by a brick wall' and determine its start and end seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/q-hQ_VNct_w_360.0_510.0.mp4
| 150
|
[
[
10,
40
]
] |
[
{
"from": "human",
"value": "<image>\nDetect and report the start and end timestamps of the video segment that semantically matches the given textual query 'Someone holding a large piece of cloth is winding it with a machine on the table'.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/q-hQ_VNct_w_210.0_360.0.mp4
| 150
|
[
[
64,
66
],
[
72,
74
],
[
82,
90
],
[
114,
120
],
[
122,
126
],
[
146,
150
]
] |
[
{
"from": "human",
"value": "<image>\nDetect and report the start and end timestamps of the video segment that semantically matches the given textual query 'Dog explores the area around a hiking trail'.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/q-hQ_VNct_w_60.0_210.0.mp4
| 150
|
[
[
42,
64
]
] |
[
{
"from": "human",
"value": "<image>\nLocalize the visual content described by the given textual query 'Dog is lying in a chair' in the video, and output the start and end timestamps in seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/q-hQ_VNct_w_660.0_810.0.mp4
| 150
|
[
[
20,
84
]
] |
[
{
"from": "human",
"value": "<image>\nFind the video segment that corresponds to the given textual query 'A lady is holding her scratch letters and showing the board to talk through what to do' and determine its start and end seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/q-hQ_VNct_w_510.0_660.0.mp4
| 150
|
[
[
118,
150
]
] |
[
{
"from": "human",
"value": "<image>\nFind the video segment that corresponds to the given textual query 'a woman in a white sweater addresses the camera' and determine its start and end seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/zjlFWTeOT3c_60.0_210.0.mp4
| 150
|
[
[
0,
56
],
[
78,
80
]
] |
[
{
"from": "human",
"value": "<image>\nFind the video segment that corresponds to the given textual query 'Clashes erupts at BML protest in Dallas' and determine its start and end seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/RUcLuQ17UV8_60.0_210.0.mp4
| 150
|
[
[
4,
10
],
[
12,
26
],
[
68,
92
],
[
112,
120
],
[
124,
126
],
[
140,
144
]
] |
[
{
"from": "human",
"value": "<image>\nLocalize the visual content described by the given textual query 'Justine Bieber giving TV interview' in the video, and output the start and end timestamps in seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/RUcLuQ17UV8_360.0_510.0.mp4
| 150
|
[
[
36,
46
],
[
48,
56
],
[
60,
62
],
[
74,
78
],
[
82,
86
],
[
96,
100
],
[
134,
150
]
] |
[
{
"from": "human",
"value": "<image>\nThe given natural language query 'Three people are hanging out in a modernist living room' is semantically aligned with a video moment, please give the start time and end time of the video moment.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/RUcLuQ17UV8_210.0_360.0.mp4
| 150
|
[
[
4,
20
]
] |
[
{
"from": "human",
"value": "<image>\nLocate the visual content mentioned in the text query 'Bride and groom walk down the aisle and kiss' within the video using timestamps.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/RUcLuQ17UV8_510.0_660.0.mp4
| 150
|
[
[
80,
108
],
[
110,
112
],
[
116,
120
]
] |
[
{
"from": "human",
"value": "<image>\nLocate the visual content mentioned in the text query 'Blonde man is wearing a happy face shirt' within the video using timestamps.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/RUcLuQ17UV8_660.0_810.0.mp4
| 150
|
[
[
72,
82
],
[
84,
116
]
] |
[
{
"from": "human",
"value": "<image>\nLocalize the visual content described by the given textual query 'Two guys sitting on the grass outside talks about aging and experiences' in the video, and output the start and end timestamps in seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/DEQiBbkiVGM_360.0_510.0.mp4
| 150
|
[
[
56,
96
]
] |
[
{
"from": "human",
"value": "<image>\nThe given natural language query 'Friends are having a conversation in the car' is semantically aligned with a video moment, please give the start time and end time of the video moment.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/DEQiBbkiVGM_360.0_510.0.mp4
| 150
|
[
[
20,
48
]
] |
[
{
"from": "human",
"value": "<image>\nGiven a textual query: 'Two guys laughing and talking to each other inside an elevator', when does the described content occur in the video? Please return the timestamp in seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/DEQiBbkiVGM_60.0_210.0.mp4
| 150
|
[
[
68,
90
]
] |
[
{
"from": "human",
"value": "<image>\nLocalize the visual content described by the given textual query 'Man with dread does his hair routine in the mirror' in the video, and output the start and end timestamps in seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/DEQiBbkiVGM_60.0_210.0.mp4
| 150
|
[
[
68,
90
]
] |
[
{
"from": "human",
"value": "<image>\nFind the video segment that corresponds to the given textual query 'Man with dreads does his haircare routine' and determine its start and end seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/DEQiBbkiVGM_210.0_360.0.mp4
| 150
|
[
[
46,
106
]
] |
[
{
"from": "human",
"value": "<image>\nLocalize the visual content described by the given textual query 'Wiz Khalifa performs on stage as large crowd vibes to his song' in the video, and output the start and end timestamps in seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/DEQiBbkiVGM_210.0_360.0.mp4
| 150
|
[
[
0,
36
]
] |
[
{
"from": "human",
"value": "<image>\nFind the video segment that corresponds to the given textual query 'Rapper Wiz Khalifa warms up for a concert in his bus' and determine its start and end seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/DEQiBbkiVGM_660.0_810.0.mp4
| 150
|
[
[
32,
44
],
[
46,
50
],
[
54,
58
],
[
60,
62
],
[
66,
88
],
[
92,
100
],
[
102,
120
]
] |
[
{
"from": "human",
"value": "<image>\nGiven a textual query: 'Wiz goes on stage a performs a song', when does the described content occur in the video? Please return the timestamp in seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/DEQiBbkiVGM_510.0_660.0.mp4
| 150
|
[
[
54,
88
]
] |
[
{
"from": "human",
"value": "<image>\nFind the video segment that corresponds to the given textual query 'Man looks down at his phone' and determine its start and end seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/T09J6x1b-wo_360.0_510.0.mp4
| 150
|
[
[
66,
92
]
] |
[
{
"from": "human",
"value": "<image>\nLocalize the visual content described by the given textual query 'A woman on a yellow jacket talking while a slideshow of protest photos is displayed' in the video, and output the start and end timestamps in seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/T09J6x1b-wo_60.0_210.0.mp4
| 150
|
[
[
0,
4
],
[
112,
150
]
] |
[
{
"from": "human",
"value": "<image>\nFind the video segment that corresponds to the given textual query 'A news reader and a reporter discussing about the police on trial for beating' and determine its start and end seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/T09J6x1b-wo_210.0_360.0.mp4
| 150
|
[
[
0,
6
],
[
8,
12
],
[
18,
150
]
] |
[
{
"from": "human",
"value": "<image>\nFind the video segment that corresponds to the given textual query 'A investigative reporter summarising how and why a defendant admits a false statemnt' and determine its start and end seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/jHf9qqTkiJc_60.0_210.0.mp4
| 128
|
[
[
0,
56
]
] |
[
{
"from": "human",
"value": "<image>\nLocalize the visual content described by the given textual query 'Official explain measures taken after extreme weather conditions' in the video, and output the start and end timestamps in seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/H1pdw-A16RY_210.0_360.0.mp4
| 130
|
[
[
0,
118
]
] |
[
{
"from": "human",
"value": "<image>\nThe given natural language query 'A woman with long straight brown hair sits in a white chair and talks' is semantically aligned with a video moment, please give the start time and end time of the video moment.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/H1pdw-A16RY_60.0_210.0.mp4
| 150
|
[
[
0,
110
],
[
138,
150
]
] |
[
{
"from": "human",
"value": "<image>\nLocalize the visual content described by the given textual query 'Woman in black top talks to the camera about her vlogging group' in the video, and output the start and end timestamps in seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/8FTpvblSAA0_360.0_510.0.mp4
| 150
|
[
[
126,
150
]
] |
[
{
"from": "human",
"value": "<image>\nDetect and report the start and end timestamps of the video segment that semantically matches the given textual query 'Man and woman have a conversation in their kitchen'.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/8FTpvblSAA0_60.0_210.0.mp4
| 150
|
[
[
130,
144
]
] |
[
{
"from": "human",
"value": "<image>\nGiven a textual query: 'A baby is standing up in it's crib', when does the described content occur in the video? Please return the timestamp in seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/8FTpvblSAA0_210.0_360.0.mp4
| 150
|
[
[
80,
128
]
] |
[
{
"from": "human",
"value": "<image>\nFind the video segment that corresponds to the given textual query 'A toddler explores a kitchen drawer while her mom is talking about nutritional supplements' and determine its start and end seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/8FTpvblSAA0_510.0_660.0.mp4
| 150
|
[
[
122,
144
]
] |
[
{
"from": "human",
"value": "<image>\nFind the video segment that corresponds to the given textual query 'The small boy in the black shirt stands there and talks with his mom' and determine its start and end seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/8FTpvblSAA0_660.0_810.0.mp4
| 150
|
[
[
40,
54
]
] |
[
{
"from": "human",
"value": "<image>\nFind the video segment that corresponds to the given textual query 'A young child runs through a kitchen into the living room and to its parents' and determine its start and end seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/4UePoM8pZjM_210.0_360.0.mp4
| 150
|
[
[
4,
8
],
[
14,
24
],
[
26,
28
],
[
38,
42
],
[
46,
48
],
[
50,
52
],
[
66,
68
],
[
76,
82
],
[
86,
88
],
[
92,
94
],
[
104,
106
],
[
112,
114
],
[
120,
122
],
[
124,
128
],
[
130,
134
],
[
140,
142
],
[
144,
146
],
[
148,
150
]
] |
[
{
"from": "human",
"value": "<image>\nLocate the visual content mentioned in the text query 'Brothers day celebration with soulmate' within the video using timestamps.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/4UePoM8pZjM_60.0_210.0.mp4
| 150
|
[
[
54,
72
],
[
76,
80
]
] |
[
{
"from": "human",
"value": "<image>\nGiven a textual query: 'A man wearing a backpack rides up an escalator', when does the described content occur in the video? Please return the timestamp in seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/4UePoM8pZjM_360.0_510.0.mp4
| 150
|
[
[
0,
2
],
[
10,
24
],
[
38,
44
]
] |
[
{
"from": "human",
"value": "<image>\nLocate the visual content mentioned in the text query 'Man in black shirt eats food' within the video using timestamps.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/qhcSDJOEXA4_360.0_510.0.mp4
| 150
|
[
[
0,
90
],
[
96,
102
],
[
108,
114
],
[
126,
150
]
] |
[
{
"from": "human",
"value": "<image>\nFind the video segment that corresponds to the given textual query 'Man and woman walk down the street together' and determine its start and end seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/qhcSDJOEXA4_210.0_360.0.mp4
| 150
|
[
[
72,
128
]
] |
[
{
"from": "human",
"value": "<image>\nFind the video segment that corresponds to the given textual query 'Nightmare before Christmas puppets are moving on a dark stage' and determine its start and end seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/qhcSDJOEXA4_60.0_210.0.mp4
| 150
|
[
[
26,
44
],
[
48,
58
],
[
64,
66
]
] |
[
{
"from": "human",
"value": "<image>\nThe given natural language query 'A woman filming Disney characters made of pumpkins at a Disney Park' is semantically aligned with a video moment, please give the start time and end time of the video moment.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/qhcSDJOEXA4_660.0_810.0.mp4
| 150
|
[
[
52,
88
]
] |
[
{
"from": "human",
"value": "<image>\nLocate the visual content mentioned in the text query 'a young girl meets jasmine' within the video using timestamps.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/qhcSDJOEXA4_510.0_660.0.mp4
| 150
|
[
[
40,
48
],
[
60,
70
]
] |
[
{
"from": "human",
"value": "<image>\nThe given natural language query 'Dad is holding his daughter in his arms' is semantically aligned with a video moment, please give the start time and end time of the video moment.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/sy4G_m_eT5k_60.0_210.0.mp4
| 150
|
[
[
90,
150
]
] |
[
{
"from": "human",
"value": "<image>\nDetect and report the start and end timestamps of the video segment that semantically matches the given textual query 'A gangster makes a blood promise in a fancy restaurant'.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/ULLVJdXSNNQ_60.0_210.0.mp4
| 150
|
[
[
74,
88
]
] |
[
{
"from": "human",
"value": "<image>\nLocalize the visual content described by the given textual query 'Overhead shots of salad in a white bowl' in the video, and output the start and end timestamps in seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/ULLVJdXSNNQ_360.0_510.0.mp4
| 150
|
[
[
52,
90
]
] |
[
{
"from": "human",
"value": "<image>\nLocalize the visual content described by the given textual query 'Food is placed down on the table and pointed at' in the video, and output the start and end timestamps in seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/ULLVJdXSNNQ_210.0_360.0.mp4
| 150
|
[
[
0,
38
]
] |
[
{
"from": "human",
"value": "<image>\nThe given natural language query 'Asian woman walks down a fenced off sidewalk' is semantically aligned with a video moment, please give the start time and end time of the video moment.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/ETEd0NdgyCg_60.0_210.0.mp4
| 150
|
[
[
0,
24
]
] |
[
{
"from": "human",
"value": "<image>\nThe given natural language query 'Weather broadcaster explains about hurricane development' is semantically aligned with a video moment, please give the start time and end time of the video moment.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/vh0ZcfGjQSA_60.0_210.0.mp4
| 150
|
[
[
50,
58
],
[
90,
98
],
[
100,
102
]
] |
[
{
"from": "human",
"value": "<image>\nFind the video segment that corresponds to the given textual query 'The dog is wearing a red jacket' and determine its start and end seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/vh0ZcfGjQSA_210.0_360.0.mp4
| 150
|
[
[
34,
150
]
] |
[
{
"from": "human",
"value": "<image>\nGiven a textual query: 'A person rides a horse in an indoor stable', when does the described content occur in the video? Please return the timestamp in seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
tvg
|
qvhighlights
|
qvh/videos/vh0ZcfGjQSA_360.0_510.0.mp4
| 150
|
[
[
0,
46
]
] |
[
{
"from": "human",
"value": "<image>\nGiven a textual query: 'Jockey is riding a brown horse around a ring', when does the described content occur in the video? Please return the timestamp in seconds.The format of your response should be: 'The event happens in <start time> - <end time>''."
},
{
"from": "gpt",
"value": "The event happens in <vid> - <vid>."
}
] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.