messages
listlengths 3
3
| videos
listlengths 1
1
| label
stringclasses 2
values | qid
stringlengths 16
21
| query
stringlengths 31
489
| pos_video_id
stringlengths 13
13
| pos_index
int64 -1
9.89k
| neg_video_id
stringlengths 0
13
| neg_index
int64 -1
9.89k
| neg_rank
int64 0
5
| neg_caption
stringlengths 31
468
⌀ |
|---|---|---|---|---|---|---|---|---|---|---|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A large pile of pinkish-brown insulation material is being installed on a metal roofing sheet, with a person's hand visible in the process.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of a person using their hands, but it is missing the specific context of installing pinkish-brown insulation material on a metal roofing sheet. Query Rewrite: A person's hand is installing a large pile of pinkish-brown insulation material on a metal roofing sheet.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_lgXtDr9pNAk.mp4"
] |
neg
|
1714394113.741532_fig
|
A large pile of pinkish-brown insulation material is being installed on a metal roofing sheet, with a person's hand visible in the process.
|
v_bTE3HEy_kVk
| 6,182
|
v_lgXtDr9pNAk
| 5,125
| 4
|
Adjusting roof flashing: the person manipulates it with hands and tools.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A large pile of pinkish-brown insulation material is being installed on a metal roofing sheet, with a person's hand visible in the process.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of people engaged in installation work, but it is missing the specific context of installing pinkish-brown insulation material on a metal roofing sheet. Query Rewrite: A person's hand is seen installing a large pile of pinkish-brown insulation material onto a metal roofing sheet.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_DG9-yffcXS4.mp4"
] |
neg
|
1714394113.741532_fig
|
A large pile of pinkish-brown insulation material is being installed on a metal roofing sheet, with a person's hand visible in the process.
|
v_bTE3HEy_kVk
| 6,182
|
v_DG9-yffcXS4
| 86
| 5
|
Amidst a snowy backdrop, a man in a yellow jacket installs new window bars with assistance from another man, as a van with a 'WIRING' sign sits idly by.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"The man, wearing a brown shirt, sits in front of a drum set with a microphone stand and speaks to the camera, wielding drumsticks and looking around at his surroundings.\""
},
{
"role": "assistant",
"content": "This video contains the moment which describes \"The man, wearing a brown shirt, sits in front of a drum set with a microphone stand and speaks to the camera, wielding drumsticks and looking around at his surroundings.\". The embedding is Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_AElVyfjSHIo.mp4"
] |
pos
|
1714394113.701246_fig
|
The man, wearing a brown shirt, sits in front of a drum set with a microphone stand and speaks to the camera, wielding drumsticks and looking around at his surroundings.
|
v_AElVyfjSHIo
| 6,948
| -1
| 0
| null |
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"The man, wearing a brown shirt, sits in front of a drum set with a microphone stand and speaks to the camera, wielding drumsticks and looking around at his surroundings.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to both men sitting in front of a drum set and holding drumsticks, but it is missing the brown shirt, the microphone stand, and the specific action of speaking to the camera while looking around at the surroundings. Query Rewrite: A man in a brown shirt sits in front of a drum set with a microphone stand, speaking to the camera while holding drumsticks and looking around.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v__jT57LdAHUs.mp4"
] |
neg
|
1714394113.701246_fig
|
The man, wearing a brown shirt, sits in front of a drum set with a microphone stand and speaks to the camera, wielding drumsticks and looking around at his surroundings.
|
v_AElVyfjSHIo
| 6,948
|
v__jT57LdAHUs
| 8,758
| 1
|
A man in a blue V-neck T-shirt, seated in front of a drum set, gestures while holding two sticks.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"The man, wearing a brown shirt, sits in front of a drum set with a microphone stand and speaks to the camera, wielding drumsticks and looking around at his surroundings.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of a man speaking to the camera, but it is missing the details of the man's brown shirt, the drum set in front of him, and the drumsticks he is wielding while looking around at his surroundings. Query Rewrite: A man in a brown shirt sits in front of a drum set, holding drumsticks, and speaks to the camera while looking around at his surroundings.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_8aYk3so14fE.mp4"
] |
neg
|
1714394113.701246_fig
|
The man, wearing a brown shirt, sits in front of a drum set with a microphone stand and speaks to the camera, wielding drumsticks and looking around at his surroundings.
|
v_AElVyfjSHIo
| 6,948
|
v_8aYk3so14fE
| 9,611
| 2
|
The man, bald and dressed in a dark blue shirt, speaks to the camera before stepping back to play the drums, with a wall and a light switch visible in the background.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"The man, wearing a brown shirt, sits in front of a drum set with a microphone stand and speaks to the camera, wielding drumsticks and looking around at his surroundings.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of a man sitting at a drum set and playing with drumsticks, but it is missing the specific details of the man's brown shirt, the microphone stand, and his interaction with the camera while looking around at his surroundings. Query Rewrite: A man in a brown shirt sits in front of a drum set with a microphone stand, speaking to the camera while wielding drumsticks and looking around at his surroundings.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_N8otQdjR96s.mp4"
] |
neg
|
1714394113.701246_fig
|
The man, wearing a brown shirt, sits in front of a drum set with a microphone stand and speaks to the camera, wielding drumsticks and looking around at his surroundings.
|
v_AElVyfjSHIo
| 6,948
|
v_N8otQdjR96s
| 8,462
| 3
|
A man in a checkered shirt and black pants sits at a drum set with a green snare drum, a blue bass drum, and a gold cymbal, playing with drumsticks in an indoor setting with a black background.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"The man, wearing a brown shirt, sits in front of a drum set with a microphone stand and speaks to the camera, wielding drumsticks and looking around at his surroundings.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of a man playing drums, but it is missing the specific details of the man's brown shirt, his interaction with the camera while speaking, and the use of drumsticks while looking around his surroundings. Query Rewrite: A man in a brown shirt sits in front of a drum set, holding drumsticks, speaking to the camera, and looking around at his surroundings.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_DhYJl8KAZN4.mp4"
] |
neg
|
1714394113.701246_fig
|
The man, wearing a brown shirt, sits in front of a drum set with a microphone stand and speaks to the camera, wielding drumsticks and looking around at his surroundings.
|
v_AElVyfjSHIo
| 6,948
|
v_DhYJl8KAZN4
| 2,589
| 4
|
A man with gray hair in a red shirt plays a golden set of drums in a room with wooden walls and a wooden floor, captured from various angles as he skillfully beats the drums.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"The man, wearing a brown shirt, sits in front of a drum set with a microphone stand and speaks to the camera, wielding drumsticks and looking around at his surroundings.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the man speaking to the camera, but it is missing the specific details of the man wearing a brown shirt, sitting in front of a drum set, and wielding drumsticks while looking around his surroundings. Query Rewrite: A man in a brown shirt sits in front of a drum set, holding drumsticks and speaking to the camera while looking around at his surroundings.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_8ZjUZbykp4U.mp4"
] |
neg
|
1714394113.701246_fig
|
The man, wearing a brown shirt, sits in front of a drum set with a microphone stand and speaks to the camera, wielding drumsticks and looking around at his surroundings.
|
v_AElVyfjSHIo
| 6,948
|
v_8ZjUZbykp4U
| 8,185
| 5
|
The camera follows the man's movements, panning around him while he speaks and reveals glimpses of a drummer, guitarist, and stage lights.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"In a festive atmosphere, the room comes alive with energetic dancing and socializing.\""
},
{
"role": "assistant",
"content": "This video contains the moment which describes \"In a festive atmosphere, the room comes alive with energetic dancing and socializing.\". The embedding is Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_4w_46rTInXI.mp4"
] |
pos
|
1714394113.705451_fig
|
In a festive atmosphere, the room comes alive with energetic dancing and socializing.
|
v_4w_46rTInXI
| 3,337
| -1
| 0
| null |
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"In a festive atmosphere, the room comes alive with energetic dancing and socializing.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of dancing and a lively atmosphere, but it is missing the emphasis on the festive atmosphere and the socializing aspect among the participants in the intended scene. Query Rewrite: In a lively and festive atmosphere, people are energetically dancing and socializing together in a vibrant room.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_eC4l8AuAmKw.mp4"
] |
neg
|
1714394113.705451_fig
|
In a festive atmosphere, the room comes alive with energetic dancing and socializing.
|
v_4w_46rTInXI
| 3,337
|
v_eC4l8AuAmKw
| 321
| 1
|
A vibrant dance floor is illuminated with colorful lights, where a man and a woman dance together in the center, surrounded by onlookers at tables.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"In a festive atmosphere, the room comes alive with energetic dancing and socializing.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of energetic dancing and socializing, but it is missing the festive atmosphere and the specific context of a more casual gathering. Query Rewrite: In a vibrant and festive atmosphere, people are joyfully dancing and socializing in a lively room filled with decorations.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_WRXSn7DyaoI.mp4"
] |
neg
|
1714394113.705451_fig
|
In a festive atmosphere, the room comes alive with energetic dancing and socializing.
|
v_4w_46rTInXI
| 3,337
|
v_WRXSn7DyaoI
| 2,916
| 2
|
A lively wedding reception is depicted with numerous people in formal attire dancing on a spacious wooden dance floor under a brightly lit room with a chandelier.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"In a festive atmosphere, the room comes alive with energetic dancing and socializing.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the energetic atmosphere created by colorful lights, but it is missing the festive atmosphere and the specific focus on dancing and socializing among individuals. Query Rewrite: In a festive atmosphere, individuals engage in energetic dancing and socializing, bringing the room to life.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_G6qOkn-K6aw.mkv"
] |
neg
|
1714394113.705451_fig
|
In a festive atmosphere, the room comes alive with energetic dancing and socializing.
|
v_4w_46rTInXI
| 3,337
|
v_G6qOkn-K6aw
| 6,743
| 3
|
In the dimly lit room, several individuals enjoy themselves as colorful lights create an energetic atmosphere.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"In a festive atmosphere, the room comes alive with energetic dancing and socializing.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of dancing and a lively atmosphere, but it is missing the festive context and the emphasis on socializing among participants. Query Rewrite: In a lively and festive atmosphere, people are joyfully dancing and socializing together in a vibrant room.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_1VmqefkNV_8.mkv"
] |
neg
|
1714394113.705451_fig
|
In a festive atmosphere, the room comes alive with energetic dancing and socializing.
|
v_4w_46rTInXI
| 3,337
|
v_1VmqefkNV_8
| 4,197
| 4
|
Amidst neon lights, a man in yellow dances with fellow participants in a dynamic setting.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"In a festive atmosphere, the room comes alive with energetic dancing and socializing.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of dancing and a lively atmosphere, but it is missing the festive context and the emphasis on energetic socializing among a broader group of people. Query Rewrite: In a festive atmosphere, a lively room is filled with energetic dancing and socializing among a diverse group of people.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_IIAg_MFuCoY.mp4"
] |
neg
|
1714394113.705451_fig
|
In a festive atmosphere, the room comes alive with energetic dancing and socializing.
|
v_4w_46rTInXI
| 3,337
|
v_IIAg_MFuCoY
| 8,705
| 5
|
A man in dark clothing and a woman in dark attire dance in sync on a dimly lit dance floor, surrounded by a sea of dancers and colorful lights in a vibrant club.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A dance team in red shirts and black pants performs a lively routine on a basketball court at night, with the court illuminated and a basketball hoop visible.\""
},
{
"role": "assistant",
"content": "This video contains the moment which describes \"A dance team in red shirts and black pants performs a lively routine on a basketball court at night, with the court illuminated and a basketball hoop visible.\". The embedding is Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_m2DOej6tPNs.mp4"
] |
pos
|
1714394113.733_fig
|
A dance team in red shirts and black pants performs a lively routine on a basketball court at night, with the court illuminated and a basketball hoop visible.
|
v_m2DOej6tPNs
| 4,327
| -1
| 0
| null |
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A dance team in red shirts and black pants performs a lively routine on a basketball court at night, with the court illuminated and a basketball hoop visible.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of a basketball court and a choreographed routine, but it is missing the specific details of a dance team wearing red shirts and black pants performing a lively routine at night with the court illuminated and a basketball hoop visible. Query Rewrite: A dance team wearing red shirts and black pants performs a lively routine on a brightly lit basketball court at night, with a basketball hoop clearly visible.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_SiBSwSu_REQ.mp4"
] |
neg
|
1714394113.733_fig
|
A dance team in red shirts and black pants performs a lively routine on a basketball court at night, with the court illuminated and a basketball hoop visible.
|
v_m2DOej6tPNs
| 4,327
|
v_SiBSwSu_REQ
| 8,409
| 1
|
A group of colorfully dressed individuals follow a choreographed routine, walking around the basketball court.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A dance team in red shirts and black pants performs a lively routine on a basketball court at night, with the court illuminated and a basketball hoop visible.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of team members in red shirts performing a routine on a sports court, but it is missing the lively performance at night, the black pants, the illuminated basketball court, and the visible basketball hoop. Query Rewrite: A dance team in red shirts and black pants performs a lively routine on an illuminated basketball court at night, with a basketball hoop visible.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_AKoregkLJgc.mp4"
] |
neg
|
1714394113.733_fig
|
A dance team in red shirts and black pants performs a lively routine on a basketball court at night, with the court illuminated and a basketball hoop visible.
|
v_m2DOej6tPNs
| 4,327
|
v_AKoregkLJgc
| 1,745
| 2
|
Four team members in red shirts and black shorts stand in a line on an indoor sports court, performing a dance or routine with outstretched arms, surrounded by seated spectators in the bleachers.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A dance team in red shirts and black pants performs a lively routine on a basketball court at night, with the court illuminated and a basketball hoop visible.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of performers in red uniforms, but it is missing the specific context of a dance team performing on a basketball court at night with a basketball hoop visible. Query Rewrite: A dance team in red shirts and black pants performs a lively routine on a brightly lit basketball court at night, with a basketball hoop clearly visible.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_ECjumI_otNw.mp4"
] |
neg
|
1714394113.733_fig
|
A dance team in red shirts and black pants performs a lively routine on a basketball court at night, with the court illuminated and a basketball hoop visible.
|
v_m2DOej6tPNs
| 4,327
|
v_ECjumI_otNw
| 1,083
| 3
|
Synchronized cheerleaders in red and white uniforms display jumps, flips, and cartwheels.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A dance team in red shirts and black pants performs a lively routine on a basketball court at night, with the court illuminated and a basketball hoop visible.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of a dance routine on a court, but it is missing the specific details of the dance team wearing red shirts and black pants, performing at night with a basketball hoop visible. Query Rewrite: A dance team wearing red shirts and black pants performs a lively routine on a brightly lit basketball court at night, with a basketball hoop clearly visible.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_diZvvLb46O8.mkv"
] |
neg
|
1714394113.733_fig
|
A dance team in red shirts and black pants performs a lively routine on a basketball court at night, with the court illuminated and a basketball hoop visible.
|
v_m2DOej6tPNs
| 4,327
|
v_diZvvLb46O8
| 1,477
| 4
|
The energetic white-and-red team performs a complex dance routine on a court, utilizing jump ropes and precise coordination.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A dance team in red shirts and black pants performs a lively routine on a basketball court at night, with the court illuminated and a basketball hoop visible.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of a group performing a routine in bright outfits, but it is missing the basketball court setting, the night-time illumination, and the specific attire of red shirts and black pants worn by the dance team. Query Rewrite: A dance team wearing red shirts and black pants performs a lively routine on a brightly lit basketball court at night, with a basketball hoop visible in the background.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_aWKZSqbOi2Q.mp4"
] |
neg
|
1714394113.733_fig
|
A dance team in red shirts and black pants performs a lively routine on a basketball court at night, with the court illuminated and a basketball hoop visible.
|
v_m2DOej6tPNs
| 4,327
|
v_aWKZSqbOi2Q
| 4,529
| 5
|
A large group of girls dressed in red and white cheerleading outfits and white shoes are standing in formation on a dark stage, executing a synchronized flip in unison and commencing their intricate routine that includes jumps, flips, hand claps, and other synchronized movements.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"Exiting the kitchen, a man engages in a handshake and conversation with another man, the negative message about him persisting on the screen.\""
},
{
"role": "assistant",
"content": "This video contains the moment which describes \"Exiting the kitchen, a man engages in a handshake and conversation with another man, the negative message about him persisting on the screen.\". The embedding is Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_7Nj1Y-fyq_c.mp4"
] |
pos
|
1714394113.722028_fig
|
Exiting the kitchen, a man engages in a handshake and conversation with another man, the negative message about him persisting on the screen.
|
v_7Nj1Y-fyq_c
| 7,781
| -1
| 0
| null |
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"Exiting the kitchen, a man engages in a handshake and conversation with another man, the negative message about him persisting on the screen.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of two men in a kitchen setting, but it is missing the handshake, conversation, and the negative message about one of the men on the screen. Query Rewrite: A man exits the kitchen and engages in a handshake and conversation with another man, while a negative message about him is displayed on the screen.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_6jxqb8FyWnQ.mp4"
] |
neg
|
1714394113.722028_fig
|
Exiting the kitchen, a man engages in a handshake and conversation with another man, the negative message about him persisting on the screen.
|
v_7Nj1Y-fyq_c
| 7,781
|
v_6jxqb8FyWnQ
| 7,877
| 1
|
In a kitchen setting, two men in shirts start an arm wrestling match, one wearing sunglasses and a black top, the other in a white shirt, with kitchen appliances and utensils nearby.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"Exiting the kitchen, a man engages in a handshake and conversation with another man, the negative message about him persisting on the screen.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of two men interacting, but it is missing the handshake, conversation, and the negative message about one of the men on the screen. Query Rewrite: A man exits the kitchen and engages in a handshake and conversation with another man, while a negative message about him is displayed on the screen.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_unLrTQt07kI.mp4"
] |
neg
|
1714394113.722028_fig
|
Exiting the kitchen, a man engages in a handshake and conversation with another man, the negative message about him persisting on the screen.
|
v_7Nj1Y-fyq_c
| 7,781
|
v_unLrTQt07kI
| 9,014
| 2
|
In a kitchen scene, a man wearing a blue striped shirt forcefully removes another man in a blue shirt from his shoulders.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"Exiting the kitchen, a man engages in a handshake and conversation with another man, the negative message about him persisting on the screen.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of two men interacting, but it is missing the context of exiting a kitchen, engaging in a handshake and conversation, and the negative message about one of the men persisting on the screen. Query Rewrite: A man exits the kitchen and engages in a handshake and conversation with another man, while a negative message about him remains visible on the screen.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_LFOlEafI35c.mp4"
] |
neg
|
1714394113.722028_fig
|
Exiting the kitchen, a man engages in a handshake and conversation with another man, the negative message about him persisting on the screen.
|
v_7Nj1Y-fyq_c
| 7,781
|
v_LFOlEafI35c
| 6,146
| 3
|
In the kitchen, two young men engage in an arm wrestling match, utilizing a standing cat chair as an impromptu table, showcasing their strength and determination.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"Exiting the kitchen, a man engages in a handshake and conversation with another man, the negative message about him persisting on the screen.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of two men, but it is missing the context of exiting the kitchen, engaging in a handshake and conversation, and the negative message about one of the men persisting on the screen. Query Rewrite: A man exits the kitchen and engages in a handshake and conversation with another man, while a negative message about him remains visible on the screen.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_E7OCzz_XuiI.mp4"
] |
neg
|
1714394113.722028_fig
|
Exiting the kitchen, a man engages in a handshake and conversation with another man, the negative message about him persisting on the screen.
|
v_7Nj1Y-fyq_c
| 7,781
|
v_E7OCzz_XuiI
| 124
| 4
|
Two men in a kitchen, wearing black and gray shirts, grab a yellow wedge of cheese from an open refrigerator with various items visible on its shelves.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"Exiting the kitchen, a man engages in a handshake and conversation with another man, the negative message about him persisting on the screen.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of two men interacting, but it is missing the specific actions of exiting a kitchen, engaging in a handshake and conversation, and the negative message about one of the men persisting on the screen. Query Rewrite: A man exits the kitchen and shakes hands with another man while they converse, with a negative message about him displayed on the screen.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_w2fsq9BOoZo.mkv"
] |
neg
|
1714394113.722028_fig
|
Exiting the kitchen, a man engages in a handshake and conversation with another man, the negative message about him persisting on the screen.
|
v_7Nj1Y-fyq_c
| 7,781
|
v_w2fsq9BOoZo
| 1,678
| 5
|
Two men laugh and play rock, paper, scissors before a crowd.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A woman strolls by a marching band, consisting of men in uniform, including red jackets, white hats, and white gloves, as they prepare to perform or march in front of a red curtain.\""
},
{
"role": "assistant",
"content": "This video contains the moment which describes \"A woman strolls by a marching band, consisting of men in uniform, including red jackets, white hats, and white gloves, as they prepare to perform or march in front of a red curtain.\". The embedding is Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_VqTyewe2R1A.mp4"
] |
pos
|
1714394113.691197_fig
|
A woman strolls by a marching band, consisting of men in uniform, including red jackets, white hats, and white gloves, as they prepare to perform or march in front of a red curtain.
|
v_VqTyewe2R1A
| 6,626
| -1
| 0
| null |
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A woman strolls by a marching band, consisting of men in uniform, including red jackets, white hats, and white gloves, as they prepare to perform or march in front of a red curtain.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of a band in red uniforms, but it is missing the specific context of a woman strolling by the band and the preparation for a performance in front of a red curtain. Query Rewrite: A woman strolls past a marching band of men in red jackets, white hats, and white gloves, as they prepare to perform in front of a red curtain.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_OjQHQptSXk0.mp4"
] |
neg
|
1714394113.691197_fig
|
A woman strolls by a marching band, consisting of men in uniform, including red jackets, white hats, and white gloves, as they prepare to perform or march in front of a red curtain.
|
v_VqTyewe2R1A
| 6,626
|
v_OjQHQptSXk0
| 2,469
| 1
|
The band, dressed in red uniforms, maintains a straight line formation while playing drums and parading down the street.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A woman strolls by a marching band, consisting of men in uniform, including red jackets, white hats, and white gloves, as they prepare to perform or march in front of a red curtain.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of a marching band in uniform, but it is missing the specific details of a woman strolling by the band and the context of preparing to perform in front of a red curtain. Query Rewrite: A woman walks past a marching band of men in red jackets, white hats, and white gloves, as they get ready to perform in front of a red curtain.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_gS4dNLjHouo.mp4"
] |
neg
|
1714394113.691197_fig
|
A woman strolls by a marching band, consisting of men in uniform, including red jackets, white hats, and white gloves, as they prepare to perform or march in front of a red curtain.
|
v_VqTyewe2R1A
| 6,626
|
v_gS4dNLjHouo
| 9,104
| 2
|
A marching band in uniform, playing drums and other instruments, parades down the street in front of a large building with columns and a flag.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A woman strolls by a marching band, consisting of men in uniform, including red jackets, white hats, and white gloves, as they prepare to perform or march in front of a red curtain.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of a marching band, but it is missing the specific details of the men in uniform wearing red jackets, white hats, and white gloves, as well as the context of preparing to perform in front of a red curtain. Query Rewrite: A woman strolls by a marching band of men in uniform, specifically wearing red jackets, white hats, and white gloves, as they prepare to perform in front of a red curtain.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_nVk5nIE-6bM.mp4"
] |
neg
|
1714394113.691197_fig
|
A woman strolls by a marching band, consisting of men in uniform, including red jackets, white hats, and white gloves, as they prepare to perform or march in front of a red curtain.
|
v_VqTyewe2R1A
| 6,626
|
v_nVk5nIE-6bM
| 2,275
| 3
|
The band, dressed in blue with white gloves, marches and plays drums in unison, entertaining a crowd.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A woman strolls by a marching band, consisting of men in uniform, including red jackets, white hats, and white gloves, as they prepare to perform or march in front of a red curtain.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of a band, but it is missing the specific details of a woman strolling by a marching band of men in uniform, including red jackets, white hats, and white gloves, as they prepare to perform in front of a red curtain. Query Rewrite: A woman walks past a marching band of men dressed in red jackets, white hats, and white gloves, preparing to perform in front of a red curtain.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_bwcgMTHFRV8.mp4"
] |
neg
|
1714394113.691197_fig
|
A woman strolls by a marching band, consisting of men in uniform, including red jackets, white hats, and white gloves, as they prepare to perform or march in front of a red curtain.
|
v_VqTyewe2R1A
| 6,626
|
v_bwcgMTHFRV8
| 1,310
| 4
|
Flag-carrying women pass by, followed by a halted band performing.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A woman strolls by a marching band, consisting of men in uniform, including red jackets, white hats, and white gloves, as they prepare to perform or march in front of a red curtain.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of a uniformed group performing, but it is missing the specific context of a woman strolling by a marching band in front of a red curtain, as well as the details of the band's attire, including red jackets, white hats, and white gloves. Query Rewrite: A woman walks past a marching band of men in red jackets, white hats, and white gloves, preparing to perform in front of a red curtain.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_xC5RVs9mXyM.mp4"
] |
neg
|
1714394113.691197_fig
|
A woman strolls by a marching band, consisting of men in uniform, including red jackets, white hats, and white gloves, as they prepare to perform or march in front of a red curtain.
|
v_VqTyewe2R1A
| 6,626
|
v_xC5RVs9mXyM
| 9,384
| 5
|
A uniformed group plays musical instruments for a crowd, who take photos and watch.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"The scene shows three children engaged in raking leaves on a grassy lawn with trees and scattered leaves. Two boys in dark clothing are raking to the right, and a little girl in a blue jacket is running behind them, both holding red rakes.\""
},
{
"role": "assistant",
"content": "This video contains the moment which describes \"The scene shows three children engaged in raking leaves on a grassy lawn with trees and scattered leaves. Two boys in dark clothing are raking to the right, and a little girl in a blue jacket is running behind them, both holding red rakes.\". The embedding is Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_I2Ctmi7C56k.mp4"
] |
pos
|
1714394113.675482_fig
|
The scene shows three children engaged in raking leaves on a grassy lawn with trees and scattered leaves. Two boys in dark clothing are raking to the right, and a little girl in a blue jacket is running behind them, both holding red rakes.
|
v_I2Ctmi7C56k
| 7,198
| -1
| 0
| null |
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"The scene shows three children engaged in raking leaves on a grassy lawn with trees and scattered leaves. Two boys in dark clothing are raking to the right, and a little girl in a blue jacket is running behind them, both holding red rakes.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of children raking leaves, but it is missing the specific details of three children, including two boys in dark clothing and a little girl in a blue jacket, as well as the action of running behind them with red rakes on a grassy lawn. Query Rewrite: The scene shows three children on a grassy lawn, with two boys in dark clothing raking leaves to the right and a little girl in a blue jacket running behind them, all holding red rakes among scattered leaves and trees.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v__I8sqYLhRKM.mp4"
] |
neg
|
1714394113.675482_fig
|
The scene shows three children engaged in raking leaves on a grassy lawn with trees and scattered leaves. Two boys in dark clothing are raking to the right, and a little girl in a blue jacket is running behind them, both holding red rakes.
|
v_I2Ctmi7C56k
| 7,198
|
v__I8sqYLhRKM
| 8,234
| 1
|
In an autumn yard, an older girl instructs a younger one on raking leaves as another girl watches from a slide.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"The scene shows three children engaged in raking leaves on a grassy lawn with trees and scattered leaves. Two boys in dark clothing are raking to the right, and a little girl in a blue jacket is running behind them, both holding red rakes.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of two boys raking leaves, but it is missing the little girl in a blue jacket running behind them and the detail of the grassy lawn with scattered leaves. Query Rewrite: The scene shows a little girl in a blue jacket running behind two boys in dark clothing who are raking leaves on a grassy lawn, with trees and scattered leaves around them, both holding red rakes.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_x0tjkH_zfXA.mp4"
] |
neg
|
1714394113.675482_fig
|
The scene shows three children engaged in raking leaves on a grassy lawn with trees and scattered leaves. Two boys in dark clothing are raking to the right, and a little girl in a blue jacket is running behind them, both holding red rakes.
|
v_I2Ctmi7C56k
| 7,198
|
v_x0tjkH_zfXA
| 2,681
| 2
|
Two boys in fall attire are raking leaves in a yard with large green and red rakes, creating a pile near a fence. One boy wears a red and black striped hat and gray jacket, while the other wears a green and black striped hoodie and blue jeans.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"The scene shows three children engaged in raking leaves on a grassy lawn with trees and scattered leaves. Two boys in dark clothing are raking to the right, and a little girl in a blue jacket is running behind them, both holding red rakes.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of a little girl and the action of raking, but it is missing the two boys in dark clothing, the specific context of raking leaves on a grassy lawn, and the detail of the red rakes being held by the children. Query Rewrite: The scene shows two boys in dark clothing raking leaves on a grassy lawn, while a little girl in a blue jacket runs behind them, all holding red rakes amidst trees and scattered leaves.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_Dty6AUdBdrE.mp4"
] |
neg
|
1714394113.675482_fig
|
The scene shows three children engaged in raking leaves on a grassy lawn with trees and scattered leaves. Two boys in dark clothing are raking to the right, and a little girl in a blue jacket is running behind them, both holding red rakes.
|
v_I2Ctmi7C56k
| 7,198
|
v_Dty6AUdBdrE
| 2,292
| 3
|
The little girl attempts to rake but is unable to move it, then gets distracted by kids, allowing the woman to take over.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"The scene shows three children engaged in raking leaves on a grassy lawn with trees and scattered leaves. Two boys in dark clothing are raking to the right, and a little girl in a blue jacket is running behind them, both holding red rakes.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of fallen leaves in a yard, but it is missing the three children, the specific actions of raking, the distinct clothing of the children, and the interaction between them. Query Rewrite: The scene shows two boys in dark clothing raking leaves to the right while a little girl in a blue jacket runs behind them, all on a grassy lawn with trees and scattered leaves.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_hYj38k-VOFM.mp4"
] |
neg
|
1714394113.675482_fig
|
The scene shows three children engaged in raking leaves on a grassy lawn with trees and scattered leaves. Two boys in dark clothing are raking to the right, and a little girl in a blue jacket is running behind them, both holding red rakes.
|
v_I2Ctmi7C56k
| 7,198
|
v_hYj38k-VOFM
| 4,937
| 4
|
A person in a blue sock stands in a yard filled with fallen leaves, raking and pausing to adjust something in their pocket.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"The scene shows three children engaged in raking leaves on a grassy lawn with trees and scattered leaves. Two boys in dark clothing are raking to the right, and a little girl in a blue jacket is running behind them, both holding red rakes.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of children raking leaves in a grassy area, but it is missing the specific details of three children, including two boys in dark clothing and a little girl in a blue jacket running behind them with red rakes. Query Rewrite: The scene shows three children on a grassy lawn with trees and scattered leaves: two boys in dark clothing raking to the right and a little girl in a blue jacket running behind them, all holding red rakes.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_h4phcAYe0aE.mp4"
] |
neg
|
1714394113.675482_fig
|
The scene shows three children engaged in raking leaves on a grassy lawn with trees and scattered leaves. Two boys in dark clothing are raking to the right, and a little girl in a blue jacket is running behind them, both holding red rakes.
|
v_I2Ctmi7C56k
| 7,198
|
v_h4phcAYe0aE
| 4,174
| 5
|
A boy wearing green rain boots and a blue and white jacket rakes leaves in a grassy area, with a woman in green jacket and blue jeans holding a rake and supervising, and a pile of leaves on the ground against a wooden fence.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"In a swimming pool with blue water, a woman wearing a black bikini supports a girl in a white and black bikini.\""
},
{
"role": "assistant",
"content": "This video contains the moment which describes \"In a swimming pool with blue water, a woman wearing a black bikini supports a girl in a white and black bikini.\". The embedding is Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_4UIAc9irBLs.mp4"
] |
pos
|
1714394113.681803_fig
|
In a swimming pool with blue water, a woman wearing a black bikini supports a girl in a white and black bikini.
|
v_4UIAc9irBLs
| 3,256
| -1
| 0
| null |
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"In a swimming pool with blue water, a woman wearing a black bikini supports a girl in a white and black bikini.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to both being set in a swimming pool environment, but it is missing the specific focus on a woman in a black bikini supporting a girl in a white and black bikini. Query Rewrite: In a swimming pool with blue water, a woman in a black bikini is supporting a girl who is wearing a white and black bikini.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_jnnHUCs-t3o.mp4"
] |
neg
|
1714394113.681803_fig
|
In a swimming pool with blue water, a woman wearing a black bikini supports a girl in a white and black bikini.
|
v_4UIAc9irBLs
| 3,256
|
v_jnnHUCs-t3o
| 3,288
| 1
|
A lively scene at a swimming pool shows several people playing a game with a ball, surrounded by chairs and tables, with a woman in a black swimsuit holding a black object nearby and palm trees in the background.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"In a swimming pool with blue water, a woman wearing a black bikini supports a girl in a white and black bikini.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to both being set in a pool environment, but it is missing the specific details of a woman in a black bikini supporting a girl in a white and black bikini. Query Rewrite: In a swimming pool with blue water, a woman in a black bikini is supporting a girl who is wearing a white and black bikini.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_qTvZtsi-3KU.mp4"
] |
neg
|
1714394113.681803_fig
|
In a swimming pool with blue water, a woman wearing a black bikini supports a girl in a white and black bikini.
|
v_4UIAc9irBLs
| 3,256
|
v_qTvZtsi-3KU
| 8,004
| 2
|
Women interact with babies in a pool setting.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"In a swimming pool with blue water, a woman wearing a black bikini supports a girl in a white and black bikini.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of a swimming pool and a woman in a black swimsuit, but it is missing the girl in a white and black bikini and the supportive interaction between the woman and the girl. Query Rewrite: In a swimming pool with blue water, a woman in a black bikini supports a girl who is wearing a white and black bikini.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_CXb3Ce7s6G0.mkv"
] |
neg
|
1714394113.681803_fig
|
In a swimming pool with blue water, a woman wearing a black bikini supports a girl in a white and black bikini.
|
v_4UIAc9irBLs
| 3,256
|
v_CXb3Ce7s6G0
| 2,970
| 3
|
In a leisurely setting by a large swimming pool with a wooden deck, a woman in a black swimsuit stands in the water, mimicking the man's actions of moving her arms and dipping her head under the water.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"In a swimming pool with blue water, a woman wearing a black bikini supports a girl in a white and black bikini.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of a woman in a black bikini at a pool, but it is missing the girl in a white and black bikini being supported by the woman and the specific focus on their interaction in the water. Query Rewrite: In a swimming pool with blue water, a woman in a black bikini is supporting a girl in a white and black bikini.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_uWnlmM2uOs4.mp4"
] |
neg
|
1714394113.681803_fig
|
In a swimming pool with blue water, a woman wearing a black bikini supports a girl in a white and black bikini.
|
v_4UIAc9irBLs
| 3,256
|
v_uWnlmM2uOs4
| 3,451
| 4
|
A woman in a black bikini receives sunscreen from her friend, who wears a bikini and holds a Coke can, both are at the poolside with other people enjoying their day.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"In a swimming pool with blue water, a woman wearing a black bikini supports a girl in a white and black bikini.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of a swimming pool, but it is missing the specific details of a woman in a black bikini supporting a girl in a white and black bikini. Query Rewrite: In a swimming pool with blue water, a woman in a black bikini is supporting a girl who is wearing a white and black bikini.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_8kNk_fzmTyI.mp4"
] |
neg
|
1714394113.681803_fig
|
In a swimming pool with blue water, a woman wearing a black bikini supports a girl in a white and black bikini.
|
v_4UIAc9irBLs
| 3,256
|
v_8kNk_fzmTyI
| 331
| 5
|
At the pool edge, she addresses the camera before the scene shifts to diverse water scenes.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"In the airport, individuals wait, walk, and sleep while carrying luggage.\""
},
{
"role": "assistant",
"content": "This video contains the moment which describes \"In the airport, individuals wait, walk, and sleep while carrying luggage.\". The embedding is Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_9V7cMp_w1_0.mp4"
] |
pos
|
1714394113.734068_fig
|
In the airport, individuals wait, walk, and sleep while carrying luggage.
|
v_9V7cMp_w1_0
| 6,160
| -1
| 0
| null |
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"In the airport, individuals wait, walk, and sleep while carrying luggage.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of individuals carrying bags, but it is missing the airport setting and the actions of waiting, walking, and sleeping. Query Rewrite: In the airport, individuals wait and walk while carrying luggage, with some resting or sleeping in the terminal.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_HVD8N0bbPVo.mkv"
] |
neg
|
1714394113.734068_fig
|
In the airport, individuals wait, walk, and sleep while carrying luggage.
|
v_9V7cMp_w1_0
| 6,160
|
v_HVD8N0bbPVo
| 5,071
| 1
|
Inside an elevator with wooden paneling, a man in a pink shirt acts erratically, punching the air while carrying a bag, with three other individuals present.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"In the airport, individuals wait, walk, and sleep while carrying luggage.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of individuals sleeping, but it is missing the context of an airport setting and the activities of waiting and walking while carrying luggage. Query Rewrite: In the airport, individuals wait and walk while carrying luggage, with some resting or sleeping nearby.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_zQCqzIgIpHI.mp4"
] |
neg
|
1714394113.734068_fig
|
In the airport, individuals wait, walk, and sleep while carrying luggage.
|
v_9V7cMp_w1_0
| 6,160
|
v_zQCqzIgIpHI
| 3,926
| 2
|
Several people are sleeping in the dark quarters, their forms barely distinguishable in the darkness.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"In the airport, individuals wait, walk, and sleep while carrying luggage.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of individuals moving through a public space, but it is missing the specific context of an airport and the activities of waiting, walking, and sleeping while carrying luggage. Query Rewrite: In the airport, individuals carry luggage while waiting, walking, and sleeping.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_thvpt_lUxTI.mp4"
] |
neg
|
1714394113.734068_fig
|
In the airport, individuals wait, walk, and sleep while carrying luggage.
|
v_9V7cMp_w1_0
| 6,160
|
v_thvpt_lUxTI
| 4,054
| 3
|
Amidst snowy urban scene, pedestrians navigate with care, varying strides, and forks and cars in motion.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"In the airport, individuals wait, walk, and sleep while carrying luggage.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of individuals handling items, but it is missing the context of an airport setting and the activities of waiting, walking, and sleeping while carrying luggage. Query Rewrite: In the airport, individuals are seen waiting, walking, and sleeping while they carry their luggage.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_OUfVZuWyqJQ.mp4"
] |
neg
|
1714394113.734068_fig
|
In the airport, individuals wait, walk, and sleep while carrying luggage.
|
v_9V7cMp_w1_0
| 6,160
|
v_OUfVZuWyqJQ
| 6,545
| 4
|
On the bed, the man opens the bag and spreads out his cash for counting.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"In the airport, individuals wait, walk, and sleep while carrying luggage.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of individuals engaged in various activities, but it is missing the specific context of an airport setting where people are waiting, walking, and sleeping while carrying luggage. Query Rewrite: In the airport, individuals wait and walk while carrying luggage, with some sleeping nearby.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_DzdNjXensv0.mp4"
] |
neg
|
1714394113.734068_fig
|
In the airport, individuals wait, walk, and sleep while carrying luggage.
|
v_9V7cMp_w1_0
| 6,160
|
v_DzdNjXensv0
| 9,629
| 5
|
A bustling marketplace filled with diverse activities, including shopping, shoe shining, and children accompanying parents.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"An indoor event at University of Mary Washington is depicted, with a blue sign prominently displayed above a podium.\""
},
{
"role": "assistant",
"content": "This video contains the moment which describes \"An indoor event at University of Mary Washington is depicted, with a blue sign prominently displayed above a podium.\". The embedding is Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v__BWSmg81iMM.mp4"
] |
pos
|
1714394113.698765_fig
|
An indoor event at University of Mary Washington is depicted, with a blue sign prominently displayed above a podium.
|
v__BWSmg81iMM
| 7,946
| -1
| 0
| null |
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"An indoor event at University of Mary Washington is depicted, with a blue sign prominently displayed above a podium.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of an indoor event atmosphere, but it is missing the specific context of the University of Mary Washington and the blue sign above a podium. Query Rewrite: An indoor event at the University of Mary Washington features a blue sign prominently displayed above a podium.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_nyEcPLZQerc.mp4"
] |
neg
|
1714394113.698765_fig
|
An indoor event at University of Mary Washington is depicted, with a blue sign prominently displayed above a podium.
|
v__BWSmg81iMM
| 7,946
|
v_nyEcPLZQerc
| 7,178
| 1
|
Everyone expresses joy and participates in collective cheers at the College of Charleston, with a marathon adding to the energetic atmosphere.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"An indoor event at University of Mary Washington is depicted, with a blue sign prominently displayed above a podium.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of a gathering with a crowd, but it is missing the indoor setting, the specific location of the University of Mary Washington, and the blue sign above a podium that are key elements of the intended scene. Query Rewrite: An indoor event at the University of Mary Washington features a blue sign prominently displayed above a podium.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_LnEed8TnWwk.mp4"
] |
neg
|
1714394113.698765_fig
|
An indoor event at University of Mary Washington is depicted, with a blue sign prominently displayed above a podium.
|
v__BWSmg81iMM
| 7,946
|
v_LnEed8TnWwk
| 1,188
| 2
|
Girls in vibrant costumes participate in a tug of war game on a grassy field, surrounded by a crowd of onlookers.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"An indoor event at University of Mary Washington is depicted, with a blue sign prominently displayed above a podium.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of a speaker, but it is missing the indoor setting at the University of Mary Washington and the blue sign above a podium. Query Rewrite: An indoor event at the University of Mary Washington features a podium with a prominently displayed blue sign.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_srR0M_HL3Ao.mp4"
] |
neg
|
1714394113.698765_fig
|
An indoor event at University of Mary Washington is depicted, with a blue sign prominently displayed above a podium.
|
v__BWSmg81iMM
| 7,946
|
v_srR0M_HL3Ao
| 2,401
| 3
|
Lisa Loverro talks at a swimming pool, followed by a slide show of Lesotho, a man speaking, Sentebale, children, and a polo match.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"An indoor event at University of Mary Washington is depicted, with a blue sign prominently displayed above a podium.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to both being indoor events, but it is missing the specific context of the University of Mary Washington and the blue sign above a podium. Query Rewrite: An indoor event at the University of Mary Washington features a podium with a prominent blue sign displayed above it.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_YnsusYE6i_0.mp4"
] |
neg
|
1714394113.698765_fig
|
An indoor event at University of Mary Washington is depicted, with a blue sign prominently displayed above a podium.
|
v__BWSmg81iMM
| 7,946
|
v_YnsusYE6i_0
| 5,958
| 4
|
Amidst tables and wine glasses at an indoor event, a woman in a blue dress takes a standing ovation from the crowd after her dance performance.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"An indoor event at University of Mary Washington is depicted, with a blue sign prominently displayed above a podium.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of a performance and a crowd, but it is missing the indoor event setting at the University of Mary Washington and the blue sign above a podium. Query Rewrite: An indoor event at the University of Mary Washington features a podium with a prominently displayed blue sign.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_ie9PgZt9svA.mp4"
] |
neg
|
1714394113.698765_fig
|
An indoor event at University of Mary Washington is depicted, with a blue sign prominently displayed above a podium.
|
v__BWSmg81iMM
| 7,946
|
v_ie9PgZt9svA
| 2,991
| 5
|
A dynamic performance by a rope jumping team of young girls in blue outfits takes place on a stage in a packed music hall, with musical instruments and a crowd of engaged spectators.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"In a field, the woman skillfully manipulates a baton with various twirling techniques and body movements.\""
},
{
"role": "assistant",
"content": "This video contains the moment which describes \"In a field, the woman skillfully manipulates a baton with various twirling techniques and body movements.\". The embedding is Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_19YCgLDhfoE.mp4"
] |
pos
|
1714394113.716109_fig
|
In a field, the woman skillfully manipulates a baton with various twirling techniques and body movements.
|
v_19YCgLDhfoE
| 1,378
| -1
| 0
| null |
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"In a field, the woman skillfully manipulates a baton with various twirling techniques and body movements.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of a woman twirling a baton, but it is missing the context of being in a field and the emphasis on skillful manipulation with various techniques and body movements. Query Rewrite: In a field, a woman skillfully twirls a baton, showcasing various techniques and body movements.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_Mk9n7isIyBs.mp4"
] |
neg
|
1714394113.716109_fig
|
In a field, the woman skillfully manipulates a baton with various twirling techniques and body movements.
|
v_19YCgLDhfoE
| 1,378
|
v_Mk9n7isIyBs
| 1,081
| 1
|
Twirling a baton on a sidewalk, she gracefully changes direction and speed, creating an engaging performance.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"In a field, the woman skillfully manipulates a baton with various twirling techniques and body movements.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the woman twirling a baton, but it is missing the specific setting of a field and the emphasis on skillful manipulation with various techniques and body movements. Query Rewrite: In a field, the woman skillfully twirls a baton using various techniques and body movements.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_mxIsgAmJHsM.mp4"
] |
neg
|
1714394113.716109_fig
|
In a field, the woman skillfully manipulates a baton with various twirling techniques and body movements.
|
v_19YCgLDhfoE
| 1,378
|
v_mxIsgAmJHsM
| 2,647
| 2
|
In an empty lot, the woman twirls a baton around her body and throws it high, catching it before spinning again.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"In a field, the woman skillfully manipulates a baton with various twirling techniques and body movements.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the woman spinning a baton in a grassy area, but it is missing the emphasis on her skillful manipulation of the baton with various twirling techniques and body movements in a field setting. Query Rewrite: In a field, a woman skillfully performs various twirling techniques with a baton, showcasing her body movements and manipulation skills.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_7iPb7mLURVc.mp4"
] |
neg
|
1714394113.716109_fig
|
In a field, the woman skillfully manipulates a baton with various twirling techniques and body movements.
|
v_19YCgLDhfoE
| 1,378
|
v_7iPb7mLURVc
| 6,501
| 3
|
A woman wearing a black top spins a baton in a grassy area with trees, showcasing her twirling skills and technique, as she continues to rotate the baton around her fingers.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"In a field, the woman skillfully manipulates a baton with various twirling techniques and body movements.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of a woman manipulating a baton, but it is missing the outdoor field setting and the emphasis on skillful twirling techniques and body movements without any mistakes. Query Rewrite: In an outdoor field, a woman skillfully twirls a baton with various techniques and fluid body movements.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_Snw8Jf1DQwg.mp4"
] |
neg
|
1714394113.716109_fig
|
In a field, the woman skillfully manipulates a baton with various twirling techniques and body movements.
|
v_19YCgLDhfoE
| 1,378
|
v_Snw8Jf1DQwg
| 87
| 4
|
In an indoor gym setting, a woman wearing a blue outfit executes a dance routine with a baton, momentarily dropping it twice before completing her performance.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"In a field, the woman skillfully manipulates a baton with various twirling techniques and body movements.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of a woman holding a baton, but it is missing the context of being in a field and the skillful manipulation of the baton with various twirling techniques and body movements. Query Rewrite: In a field, a woman skillfully twirls a baton using various techniques and body movements.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_fFu37ig0pkY.mp4"
] |
neg
|
1714394113.716109_fig
|
In a field, the woman skillfully manipulates a baton with various twirling techniques and body movements.
|
v_19YCgLDhfoE
| 1,378
|
v_fFu37ig0pkY
| 936
| 5
|
A woman in a black dress, holding a baton, twirls it as she runs along a sidewalk in front of a building.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A woman in red holds a microphone, engaging a seated girl in a lively conversation on stage.\""
},
{
"role": "assistant",
"content": "This video contains the moment which describes \"A woman in red holds a microphone, engaging a seated girl in a lively conversation on stage.\". The embedding is Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_dQs2-z3TIes.mp4"
] |
pos
|
1714394113.706193_fig
|
A woman in red holds a microphone, engaging a seated girl in a lively conversation on stage.
|
v_dQs2-z3TIes
| 864
| -1
| 0
| null |
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A woman in red holds a microphone, engaging a seated girl in a lively conversation on stage.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of a woman in red on stage, but it is missing the microphone, the lively conversation with a seated girl, and the absence of any crowd or off-camera engagement. Query Rewrite: A woman in red holds a microphone on stage, engaging in a lively conversation with a seated girl.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_qcYRPEEitZU.mp4"
] |
neg
|
1714394113.706193_fig
|
A woman in red holds a microphone, engaging a seated girl in a lively conversation on stage.
|
v_dQs2-z3TIes
| 864
|
v_qcYRPEEitZU
| 2,348
| 1
|
A woman in a red and white dress sits on stage, engaging with someone off-camera, while holding shaving cream and a razor amidst a large, observing crowd.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A woman in red holds a microphone, engaging a seated girl in a lively conversation on stage.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to both being set on a stage with lively activity, but it is missing the specific interaction between a woman in red holding a microphone and a seated girl engaged in conversation. Query Rewrite: A woman in red holds a microphone and engages a seated girl in a lively conversation on stage.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_YFTQImGp_nY.mkv"
] |
neg
|
1714394113.706193_fig
|
A woman in red holds a microphone, engaging a seated girl in a lively conversation on stage.
|
v_dQs2-z3TIes
| 864
|
v_YFTQImGp_nY
| 522
| 2
|
A lively performance on a stage: Older group members hold poles for children to swing from, surrounded by a backdrop of colorful lighting and a large circular design.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A woman in red holds a microphone, engaging a seated girl in a lively conversation on stage.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of a woman on stage with a microphone, but it is missing the interaction with a seated girl and the lively conversation aspect. Query Rewrite: A woman in red engages a seated girl in a lively conversation on stage while holding a microphone.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_8ltNqT5N-xs.mp4"
] |
neg
|
1714394113.706193_fig
|
A woman in red holds a microphone, engaging a seated girl in a lively conversation on stage.
|
v_dQs2-z3TIes
| 864
|
v_8ltNqT5N-xs
| 1,445
| 3
|
A woman in a colorful dress performs on stage, singing into a microphone and dancing energetically, ending her routine by flipping back her long hair.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A woman in red holds a microphone, engaging a seated girl in a lively conversation on stage.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of a woman in red, but it is missing the microphone, the lively conversation with a seated girl, and the stage setting. Query Rewrite: On stage, a woman in red holds a microphone while engaging in a lively conversation with a seated girl.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_SbX8KDWMhio.mp4"
] |
neg
|
1714394113.706193_fig
|
A woman in red holds a microphone, engaging a seated girl in a lively conversation on stage.
|
v_dQs2-z3TIes
| 864
|
v_SbX8KDWMhio
| 6,155
| 4
|
With a hookah hose in hand, a woman in jeans and a red shirt dances with expressive arm movements in front of a television.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A woman in red holds a microphone, engaging a seated girl in a lively conversation on stage.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of a woman in red and a microphone, but it is missing the interaction with a seated girl and the lively conversation on stage. Query Rewrite: A woman in a red outfit engages a seated girl in a lively conversation on stage while holding a microphone.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_t_D9MYkEPEo.mkv"
] |
neg
|
1714394113.706193_fig
|
A woman in red holds a microphone, engaging a seated girl in a lively conversation on stage.
|
v_dQs2-z3TIes
| 864
|
v_t_D9MYkEPEo
| 5,621
| 5
|
A woman in a vibrant red dress spins and twirls gracefully behind a microphone on a blue background, while a man in a dark suit and holding a microphone stands nearby.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A gymnast-like figure does a handstand and a rope flip in the midst of an indoor basketball court, surrounded by flags and banners, showcasing his agility and acrobatic skills.\""
},
{
"role": "assistant",
"content": "This video contains the moment which describes \"A gymnast-like figure does a handstand and a rope flip in the midst of an indoor basketball court, surrounded by flags and banners, showcasing his agility and acrobatic skills.\". The embedding is Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_cWU5059m_1Q.mp4"
] |
pos
|
1714394113.688481_fig
|
A gymnast-like figure does a handstand and a rope flip in the midst of an indoor basketball court, surrounded by flags and banners, showcasing his agility and acrobatic skills.
|
v_cWU5059m_1Q
| 342
| -1
| 0
| null |
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A gymnast-like figure does a handstand and a rope flip in the midst of an indoor basketball court, surrounded by flags and banners, showcasing his agility and acrobatic skills.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of a handstand artist performing in a gymnasium, but it is missing the specific context of an indoor basketball court, the rope flip action, and the surrounding flags and banners that highlight the gymnast's agility and acrobatic skills. Query Rewrite: A gymnast-like figure performs a handstand and a rope flip on an indoor basketball court, surrounded by flags and banners, demonstrating exceptional agility and acrobatic skills.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_Zhfm2dWtnpc.mp4"
] |
neg
|
1714394113.688481_fig
|
A gymnast-like figure does a handstand and a rope flip in the midst of an indoor basketball court, surrounded by flags and banners, showcasing his agility and acrobatic skills.
|
v_cWU5059m_1Q
| 342
|
v_Zhfm2dWtnpc
| 65
| 1
|
A handstand artist in a white shirt and black shorts showcases his skills on the wooden floor of a bustling gymnasium, as spectators watch from the sidelines and some individuals crawl around on the floor.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A gymnast-like figure does a handstand and a rope flip in the midst of an indoor basketball court, surrounded by flags and banners, showcasing his agility and acrobatic skills.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of acrobatic movements on a gym floor, but it is missing the specific context of an indoor basketball court, the gymnast-like figure, and the surrounding flags and banners that highlight the intended scene's unique setting and character. Query Rewrite: A gymnast-like figure performs a handstand and a rope flip on an indoor basketball court, surrounded by flags and banners, demonstrating exceptional agility and acrobatic skills.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_Z6xXj6tbMCo.mp4"
] |
neg
|
1714394113.688481_fig
|
A gymnast-like figure does a handstand and a rope flip in the midst of an indoor basketball court, surrounded by flags and banners, showcasing his agility and acrobatic skills.
|
v_cWU5059m_1Q
| 342
|
v_Z6xXj6tbMCo
| 7,191
| 2
|
A man in a white shirt and black shorts performs acrobatic flips and tricks on a blue gym floor, surrounded by other individuals in athletic attire practicing or observing.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A gymnast-like figure does a handstand and a rope flip in the midst of an indoor basketball court, surrounded by flags and banners, showcasing his agility and acrobatic skills.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of a young athlete performing acrobatic movements in an indoor setting, but it is missing the specific action of a rope flip and the distinct context of a basketball court. Query Rewrite: A gymnast-like figure performs a handstand and a rope flip on an indoor basketball court, surrounded by flags and banners, demonstrating exceptional agility and acrobatic skills.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_TF591ByrGYk.mp4"
] |
neg
|
1714394113.688481_fig
|
A gymnast-like figure does a handstand and a rope flip in the midst of an indoor basketball court, surrounded by flags and banners, showcasing his agility and acrobatic skills.
|
v_cWU5059m_1Q
| 342
|
v_TF591ByrGYk
| 8,946
| 3
|
A young athlete in a white shirt and black shorts practices handstands and swings on parallel bars in an indoor gymnasium, in front of a cheering crowd and various gymnastics equipment including flags and banners.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A gymnast-like figure does a handstand and a rope flip in the midst of an indoor basketball court, surrounded by flags and banners, showcasing his agility and acrobatic skills.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of a gymnast performing acrobatic moves in an indoor sports arena, but it is missing the specific details of a handstand and rope flip, as well as the context of being surrounded by flags and banners. Query Rewrite: A gymnast-like figure performs a handstand and a rope flip on an indoor basketball court, surrounded by flags and banners, demonstrating agility and acrobatic skills.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_U6M5hbiHn5Q.mp4"
] |
neg
|
1714394113.688481_fig
|
A gymnast-like figure does a handstand and a rope flip in the midst of an indoor basketball court, surrounded by flags and banners, showcasing his agility and acrobatic skills.
|
v_cWU5059m_1Q
| 342
|
v_U6M5hbiHn5Q
| 794
| 4
|
A gymnast executes various acrobatic moves in an indoor sports arena with red seats and a blue floor, showcasing their athleticism and precision in front of a cheering audience.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A gymnast-like figure does a handstand and a rope flip in the midst of an indoor basketball court, surrounded by flags and banners, showcasing his agility and acrobatic skills.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of a gymnast performing handstands, but it is missing the indoor basketball court setting and the specific action of a rope flip surrounded by flags and banners that showcase agility and acrobatic skills. Query Rewrite: A gymnast-like figure performs a handstand and a rope flip on an indoor basketball court, surrounded by flags and banners, demonstrating exceptional agility and acrobatic skills.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_W06Ojnv-Wxg.mp4"
] |
neg
|
1714394113.688481_fig
|
A gymnast-like figure does a handstand and a rope flip in the midst of an indoor basketball court, surrounded by flags and banners, showcasing his agility and acrobatic skills.
|
v_cWU5059m_1Q
| 342
|
v_W06Ojnv-Wxg
| 6,636
| 5
|
A gymnast executes a complex uneven bars routine, transitioning between bars and performing handstands and releases.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A young Black male, serving as the Public Relations representative for triple A, stands in a snowy environment and demonstrates the extent of the snow on two parked cars - a white sedan and a darker-colored sedan, both covered in a thick layer of snow.\""
},
{
"role": "assistant",
"content": "This video contains the moment which describes \"A young Black male, serving as the Public Relations representative for triple A, stands in a snowy environment and demonstrates the extent of the snow on two parked cars - a white sedan and a darker-colored sedan, both covered in a thick layer of snow.\". The embedding is Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_3J_-uEjb8UE.mp4"
] |
pos
|
1714394113.739507_fig
|
A young Black male, serving as the Public Relations representative for triple A, stands in a snowy environment and demonstrates the extent of the snow on two parked cars - a white sedan and a darker-colored sedan, both covered in a thick layer of snow.
|
v_3J_-uEjb8UE
| 1,825
| -1
| 0
| null |
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A young Black male, serving as the Public Relations representative for triple A, stands in a snowy environment and demonstrates the extent of the snow on two parked cars - a white sedan and a darker-colored sedan, both covered in a thick layer of snow.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the snowy environment and the presence of a white sedan, but it is missing the young Black male Public Relations representative, the demonstration of snow extent on two parked cars, and the darker-colored sedan covered in snow. Query Rewrite: A young Black male serves as the Public Relations representative for triple A, standing in a snowy environment and demonstrating the extent of snow covering two parked cars, a white sedan and a darker-colored sedan.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_H8f7pTaBFKk.mp4"
] |
neg
|
1714394113.739507_fig
|
A young Black male, serving as the Public Relations representative for triple A, stands in a snowy environment and demonstrates the extent of the snow on two parked cars - a white sedan and a darker-colored sedan, both covered in a thick layer of snow.
|
v_3J_-uEjb8UE
| 1,825
|
v_H8f7pTaBFKk
| 4,061
| 1
|
Amidst a backdrop of a snowy street and a building, a man in a dark jacket and hat scrapes off snow from a white sedan, while speaking to the camera and wearing a scraper in hand.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A young Black male, serving as the Public Relations representative for triple A, stands in a snowy environment and demonstrates the extent of the snow on two parked cars - a white sedan and a darker-colored sedan, both covered in a thick layer of snow.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the snowy environment and the presence of parked cars, but it is missing the specific details of a young Black male serving as the Public Relations representative for triple A and the two distinct sedans, one white and one darker-colored, both covered in a thick layer of snow. Query Rewrite: A young Black male, representing triple A, stands in a snowy environment next to two parked sedans, one white and one darker-colored, both completely covered in a thick layer of snow.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_VWmSZPIoBlw.mp4"
] |
neg
|
1714394113.739507_fig
|
A young Black male, serving as the Public Relations representative for triple A, stands in a snowy environment and demonstrates the extent of the snow on two parked cars - a white sedan and a darker-colored sedan, both covered in a thick layer of snow.
|
v_3J_-uEjb8UE
| 1,825
|
v_VWmSZPIoBlw
| 8,440
| 2
|
Amidst a snowy urban scene, a man in a black coat is shown standing on the sidewalk, with a white Smart car parked nearby, its exterior blanketed in recent snowfall.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A young Black male, serving as the Public Relations representative for triple A, stands in a snowy environment and demonstrates the extent of the snow on two parked cars - a white sedan and a darker-colored sedan, both covered in a thick layer of snow.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the snowy environment and the presence of parked cars, but it is missing the specific focus on a young Black male Public Relations representative demonstrating the extent of snow on two parked cars, one being a white sedan and the other a darker-colored sedan. Query Rewrite: A young Black male serves as the Public Relations representative for triple A, standing in a snowy environment and demonstrating the thickness of snow covering two parked cars: a white sedan and a darker-colored sedan.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_YtgiDWEY_1A.mp4"
] |
neg
|
1714394113.739507_fig
|
A young Black male, serving as the Public Relations representative for triple A, stands in a snowy environment and demonstrates the extent of the snow on two parked cars - a white sedan and a darker-colored sedan, both covered in a thick layer of snow.
|
v_3J_-uEjb8UE
| 1,825
|
v_YtgiDWEY_1A
| 6,117
| 3
|
Albertinho, wearing a dark jacket, walks through the snowy parking lot, talking into a camera while his wife helps clear snow from their cars and those nearby.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A young Black male, serving as the Public Relations representative for triple A, stands in a snowy environment and demonstrates the extent of the snow on two parked cars - a white sedan and a darker-colored sedan, both covered in a thick layer of snow.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the snowy environment, but it is missing the young Black male Public Relations representative, the demonstration of snow on two parked cars, and the specific details about the cars being a white sedan and a darker-colored sedan covered in snow. Query Rewrite: A young Black male serves as the Public Relations representative for triple A, standing in a snowy environment next to two parked cars, a white sedan and a darker-colored sedan, both covered in a thick layer of snow.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_WV_MFi9EZeQ.mp4"
] |
neg
|
1714394113.739507_fig
|
A young Black male, serving as the Public Relations representative for triple A, stands in a snowy environment and demonstrates the extent of the snow on two parked cars - a white sedan and a darker-colored sedan, both covered in a thick layer of snow.
|
v_3J_-uEjb8UE
| 1,825
|
v_WV_MFi9EZeQ
| 3,566
| 4
|
Amidst a snowy backdrop, a man in a red jacket and black hat uses a scraper to remove ice from a car's windows while addressing the camera.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A young Black male, serving as the Public Relations representative for triple A, stands in a snowy environment and demonstrates the extent of the snow on two parked cars - a white sedan and a darker-colored sedan, both covered in a thick layer of snow.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of a car covered in ice, but it is missing the young Black male Public Relations representative, the snowy environment, and the specific mention of two parked cars, one being a white sedan and the other a darker-colored sedan. Query Rewrite: A young Black male serves as the Public Relations representative for triple A, standing in a snowy environment next to two parked cars, a white sedan and a darker-colored sedan, both covered in a thick layer of snow.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_rzomS2gZGQ4.mp4"
] |
neg
|
1714394113.739507_fig
|
A young Black male, serving as the Public Relations representative for triple A, stands in a snowy environment and demonstrates the extent of the snow on two parked cars - a white sedan and a darker-colored sedan, both covered in a thick layer of snow.
|
v_3J_-uEjb8UE
| 1,825
|
v_rzomS2gZGQ4
| 8,235
| 5
|
At night, a man stands by his silver sedan with a white roof, which has a thick layer of ice on its hood and roof, parked on a street with other cars.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A man, presumably in his middle age with dark hair, applies a blue cream or gel from a tube to his chin in the comfort of a well-lit room.\""
},
{
"role": "assistant",
"content": "This video contains the moment which describes \"A man, presumably in his middle age with dark hair, applies a blue cream or gel from a tube to his chin in the comfort of a well-lit room.\". The embedding is Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_nFfMY4CQq50.mp4"
] |
pos
|
1714394113.718896_fig
|
A man, presumably in his middle age with dark hair, applies a blue cream or gel from a tube to his chin in the comfort of a well-lit room.
|
v_nFfMY4CQq50
| 530
| -1
| 0
| null |
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A man, presumably in his middle age with dark hair, applies a blue cream or gel from a tube to his chin in the comfort of a well-lit room.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the act of applying a product to the face, but it is missing the specific context of a well-lit indoor setting, the use of a blue cream or gel from a tube, and the description of the man as being in his middle age with dark hair. Query Rewrite: In a well-lit indoor room, a middle-aged man with dark hair applies a blue cream or gel from a tube to his chin.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_arTumJHoLB0.mp4"
] |
neg
|
1714394113.718896_fig
|
A man, presumably in his middle age with dark hair, applies a blue cream or gel from a tube to his chin in the comfort of a well-lit room.
|
v_nFfMY4CQq50
| 530
|
v_arTumJHoLB0
| 3,452
| 1
|
A man, wearing a sleeveless shirt and a watch, applies lotion to his face while standing outdoors, with a trail of lotion visible on his skin and trees in the background.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A man, presumably in his middle age with dark hair, applies a blue cream or gel from a tube to his chin in the comfort of a well-lit room.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of a man with dark hair applying a gel, but it is missing the specific context of a well-lit room and the action of applying the gel to his chin rather than his hands. Query Rewrite: In a well-lit room, a middle-aged man with dark hair applies a blue cream or gel from a tube to his chin.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_ueDoskgg2yw.mp4"
] |
neg
|
1714394113.718896_fig
|
A man, presumably in his middle age with dark hair, applies a blue cream or gel from a tube to his chin in the comfort of a well-lit room.
|
v_nFfMY4CQq50
| 530
|
v_ueDoskgg2yw
| 419
| 2
|
A man in a bathroom, wearing dark hair, applies a small amount of blue and white gel to his hands, likely for protection, in front of a mirror with blue tiles.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A man, presumably in his middle age with dark hair, applies a blue cream or gel from a tube to his chin in the comfort of a well-lit room.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of a man applying a substance to his body, but it is missing the specific context of a well-lit room, the application of a blue cream or gel to his chin, and the man's middle age with dark hair. Query Rewrite: In a well-lit room, a middle-aged man with dark hair applies a blue cream or gel from a tube to his chin.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_bDiwuABU45I.mp4"
] |
neg
|
1714394113.718896_fig
|
A man, presumably in his middle age with dark hair, applies a blue cream or gel from a tube to his chin in the comfort of a well-lit room.
|
v_nFfMY4CQq50
| 530
|
v_bDiwuABU45I
| 163
| 3
|
A man stands on a sandy beach, wearing blue swim trunks and holding a sunscreen bottle, as he rubs the lotion on his back, with a sandy background and a person lying on a towel in the distance.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A man, presumably in his middle age with dark hair, applies a blue cream or gel from a tube to his chin in the comfort of a well-lit room.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of a man applying a substance, but it is missing the specific context of a well-lit room, the application of blue cream or gel to his chin, and the man's dark hair and middle age appearance. Query Rewrite: In a well-lit room, a middle-aged man with dark hair applies blue cream or gel from a tube to his chin.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_SlogVhfmTVs.mp4"
] |
neg
|
1714394113.718896_fig
|
A man, presumably in his middle age with dark hair, applies a blue cream or gel from a tube to his chin in the comfort of a well-lit room.
|
v_nFfMY4CQq50
| 530
|
v_SlogVhfmTVs
| 3,707
| 4
|
In a bathroom setting, a man in a striped shirt and a watch applies liquid from a medicine dropper to a razor, showing the process to the camera.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A man, presumably in his middle age with dark hair, applies a blue cream or gel from a tube to his chin in the comfort of a well-lit room.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of a man applying cream to his face, but it is missing the specific context of a well-lit room and the detail that he is applying a blue cream or gel to his chin rather than shaving. Query Rewrite: In a well-lit room, a middle-aged man with dark hair applies a blue cream or gel to his chin.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_-YMpwZkNc2A.mp4"
] |
neg
|
1714394113.718896_fig
|
A man, presumably in his middle age with dark hair, applies a blue cream or gel from a tube to his chin in the comfort of a well-lit room.
|
v_nFfMY4CQq50
| 530
|
v_-YMpwZkNc2A
| 9,161
| 5
|
The man, in a bathroom, rubs cream on his face and shaves using a razor.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A man in a dark athletic outfit prepares to throw a hammer in a stadium, spinning it around several times before releasing it to fly into the hair of a person in the field.\""
},
{
"role": "assistant",
"content": "This video contains the moment which describes \"A man in a dark athletic outfit prepares to throw a hammer in a stadium, spinning it around several times before releasing it to fly into the hair of a person in the field.\". The embedding is Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_Xm23RMCpDd8.mp4"
] |
pos
|
1714394113.724992_fig
|
A man in a dark athletic outfit prepares to throw a hammer in a stadium, spinning it around several times before releasing it to fly into the hair of a person in the field.
|
v_Xm23RMCpDd8
| 2,264
| -1
| 0
| null |
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A man in a dark athletic outfit prepares to throw a hammer in a stadium, spinning it around several times before releasing it to fly into the hair of a person in the field.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of a hammer thrower in a dark outfit, but it is missing the specific action of the man preparing to throw the hammer and the detail of the hammer flying into the hair of a person in the field. Query Rewrite: A man in a dark athletic outfit spins a hammer several times in a stadium before releasing it, causing the hammer to fly into the hair of a person in the field.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_7oBW5W6Fdok.mp4"
] |
neg
|
1714394113.724992_fig
|
A man in a dark athletic outfit prepares to throw a hammer in a stadium, spinning it around several times before releasing it to fly into the hair of a person in the field.
|
v_Xm23RMCpDd8
| 2,264
|
v_7oBW5W6Fdok
| 5,056
| 1
|
A hammer thrower in action, spinning and throwing the hammer in a sporting event, wearing a dark outfit.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A man in a dark athletic outfit prepares to throw a hammer in a stadium, spinning it around several times before releasing it to fly into the hair of a person in the field.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of a male athlete preparing to throw a hammer, but it is missing the dark athletic outfit, the specific action of the hammer flying into the hair of a person in the field, and the context of a stadium setting. Query Rewrite: A man in a dark athletic outfit spins a hammer several times before releasing it, with the hammer flying into the hair of a person in the field at a stadium.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_KnHUAc20WEU.mp4"
] |
neg
|
1714394113.724992_fig
|
A man in a dark athletic outfit prepares to throw a hammer in a stadium, spinning it around several times before releasing it to fly into the hair of a person in the field.
|
v_Xm23RMCpDd8
| 2,264
|
v_KnHUAc20WEU
| 3,523
| 2
|
A male athlete in blue and white uniform is seen preparing, spinning, and celebrating his hammer throw on the large screen.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A man in a dark athletic outfit prepares to throw a hammer in a stadium, spinning it around several times before releasing it to fly into the hair of a person in the field.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the mention of hammer throwing and spinning, but it is missing the specific details of a man in a dark athletic outfit preparing to throw the hammer in a stadium and the hammer flying into the hair of a person in the field. Query Rewrite: A man in a dark athletic outfit stands in a stadium, preparing to throw a hammer by spinning it around several times before releasing it, causing the hammer to fly into the hair of a person in the field.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_w8lqIgJn09Y.mp4"
] |
neg
|
1714394113.724992_fig
|
A man in a dark athletic outfit prepares to throw a hammer in a stadium, spinning it around several times before releasing it to fly into the hair of a person in the field.
|
v_Xm23RMCpDd8
| 2,264
|
v_w8lqIgJn09Y
| 2,542
| 3
|
Athlete's hammer throw: spinning, releasing at distinct speeds.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A man in a dark athletic outfit prepares to throw a hammer in a stadium, spinning it around several times before releasing it to fly into the hair of a person in the field.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of a man in a dark athletic outfit and an athletic setting, but it is missing the action of throwing a hammer and the specific detail of the hammer flying into the hair of a person in the field. Query Rewrite: A man in a dark athletic outfit prepares to throw a hammer in a stadium, spinning it around before releasing it towards a person in the field.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_4R5KDUqBPcU.mp4"
] |
neg
|
1714394113.724992_fig
|
A man in a dark athletic outfit prepares to throw a hammer in a stadium, spinning it around several times before releasing it to fly into the hair of a person in the field.
|
v_Xm23RMCpDd8
| 2,264
|
v_4R5KDUqBPcU
| 3,250
| 4
|
Man in black athletic suit is practicing shot put on an outdoor circular platform with a net in the background, holding a shot put, and another person is visible in the background.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A man in a dark athletic outfit prepares to throw a hammer in a stadium, spinning it around several times before releasing it to fly into the hair of a person in the field.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of a man performing a hammer throw, but it is missing the specific details of the man's dark athletic outfit, the stadium setting, and the action of the hammer flying towards a person in the field. Query Rewrite: A man in a dark athletic outfit prepares to throw a hammer in a stadium, spinning it around several times before releasing it towards a person in the field.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_VLQQVNIZQqo.mp4"
] |
neg
|
1714394113.724992_fig
|
A man in a dark athletic outfit prepares to throw a hammer in a stadium, spinning it around several times before releasing it to fly into the hair of a person in the field.
|
v_Xm23RMCpDd8
| 2,264
|
v_VLQQVNIZQqo
| 1,629
| 5
|
Amidst a lively crowd in casual clothing, a man dressed in white shirt and dark shorts executes a hammer throw on a grassy field, capturing the attention of all present.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A group of swimmers in swimsuits swim towards the other side of an indoor pool with a blue floor and white walls, while yellow flags hang above the pool.\""
},
{
"role": "assistant",
"content": "This video contains the moment which describes \"A group of swimmers in swimsuits swim towards the other side of an indoor pool with a blue floor and white walls, while yellow flags hang above the pool.\". The embedding is Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_T49ZxNX1DFM.mp4"
] |
pos
|
1714394113.692217_fig
|
A group of swimmers in swimsuits swim towards the other side of an indoor pool with a blue floor and white walls, while yellow flags hang above the pool.
|
v_T49ZxNX1DFM
| 8,598
| -1
| 0
| null |
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A group of swimmers in swimsuits swim towards the other side of an indoor pool with a blue floor and white walls, while yellow flags hang above the pool.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of swimmers in swimsuits in a swimming pool, but it is missing the specific details of an indoor pool with a blue floor and white walls, as well as the yellow flags hanging above the pool. Query Rewrite: A group of swimmers in swimsuits swim towards the other side of an indoor pool featuring a blue floor and white walls, with yellow flags hanging above the pool.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_2b_hjshny54.mp4"
] |
neg
|
1714394113.692217_fig
|
A group of swimmers in swimsuits swim towards the other side of an indoor pool with a blue floor and white walls, while yellow flags hang above the pool.
|
v_T49ZxNX1DFM
| 8,598
|
v_2b_hjshny54
| 3,732
| 1
|
Swimmers in swimsuits practice or race in a clear swimming pool with multiple lanes, marked with dividers. They swim in a repetitive pattern, showcasing diverse techniques.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A group of swimmers in swimsuits swim towards the other side of an indoor pool with a blue floor and white walls, while yellow flags hang above the pool.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of swimmers in swimsuits swimming in an indoor pool, but it is missing the specific details of swimming towards the other side of the pool, the blue floor, the white walls, and the yellow flags hanging above the pool. Query Rewrite: A group of swimmers in swimsuits swim towards the far side of an indoor pool with a blue floor and white walls, while yellow flags hang above the water.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_TdNiQsZNJTU.mp4"
] |
neg
|
1714394113.692217_fig
|
A group of swimmers in swimsuits swim towards the other side of an indoor pool with a blue floor and white walls, while yellow flags hang above the pool.
|
v_T49ZxNX1DFM
| 8,598
|
v_TdNiQsZNJTU
| 198
| 2
|
A group of swimmers wearing swimsuits swim laps in an indoor pool with a clear blue water surface, lane markers, and a starting block, indicating a competitive or training swim session.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A group of swimmers in swimsuits swim towards the other side of an indoor pool with a blue floor and white walls, while yellow flags hang above the pool.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of an indoor pool and people swimming, but it is missing the specific details of swimmers in swimsuits swimming towards the other side of the pool, the blue floor, white walls, and the yellow flags hanging above the pool. Query Rewrite: A group of swimmers in swimsuits swim towards the far side of an indoor pool with a blue floor and white walls, while yellow flags hang above the pool.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_8L1xXJvKuv4.mp4"
] |
neg
|
1714394113.692217_fig
|
A group of swimmers in swimsuits swim towards the other side of an indoor pool with a blue floor and white walls, while yellow flags hang above the pool.
|
v_T49ZxNX1DFM
| 8,598
|
v_8L1xXJvKuv4
| 459
| 3
|
A lively scene of people swimming in an indoor pool is shown, with several onlookers in casual attire observing from the poolside.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A group of swimmers in swimsuits swim towards the other side of an indoor pool with a blue floor and white walls, while yellow flags hang above the pool.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of a swimming pool and flags, but it is missing the group of swimmers in swimsuits swimming towards the other side of the pool with a blue floor and white walls. Query Rewrite: A group of swimmers in swimsuits swim towards the far side of an indoor pool with a blue floor and white walls, while yellow flags hang above the pool.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_opCMfcAl7yw.mkv"
] |
neg
|
1714394113.692217_fig
|
A group of swimmers in swimsuits swim towards the other side of an indoor pool with a blue floor and white walls, while yellow flags hang above the pool.
|
v_T49ZxNX1DFM
| 8,598
|
v_opCMfcAl7yw
| 9,187
| 4
|
An underwater view of a swimming pool reveals a person swimming in the middle of a lane, surrounded by blue and yellow lane markers, with a group of flags visible in the background.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"A group of swimmers in swimsuits swim towards the other side of an indoor pool with a blue floor and white walls, while yellow flags hang above the pool.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to the presence of a blue pool, but it is missing the specific context of a group of swimmers in swimsuits swimming towards the other side of an indoor pool with white walls and yellow flags hanging above. Query Rewrite: A group of swimmers in swimsuits swim towards the far side of an indoor pool with white walls and yellow flags hanging above, set against a blue floor.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_7g3YO0-R4qk.mp4"
] |
neg
|
1714394113.692217_fig
|
A group of swimmers in swimsuits swim towards the other side of an indoor pool with a blue floor and white walls, while yellow flags hang above the pool.
|
v_T49ZxNX1DFM
| 8,598
|
v_7g3YO0-R4qk
| 7,899
| 5
|
Men swim laps in a blue pool with lane markers, occasionally jumping in from the side. The scene transitions to a black background, then back to the pool, where men continue their laps.
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"An aerobics instructor, identifiable by her black tank top and black shorts, leads a dynamic class in a gym setting, with participants following her lead and mirrors reflecting their movements, as exercise equipment is visible in the background.\""
},
{
"role": "assistant",
"content": "This video contains the moment which describes \"An aerobics instructor, identifiable by her black tank top and black shorts, leads a dynamic class in a gym setting, with participants following her lead and mirrors reflecting their movements, as exercise equipment is visible in the background.\". The embedding is Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_2WeG8QGOfcI.mp4"
] |
pos
|
1714394113.730561_fig
|
An aerobics instructor, identifiable by her black tank top and black shorts, leads a dynamic class in a gym setting, with participants following her lead and mirrors reflecting their movements, as exercise equipment is visible in the background.
|
v_2WeG8QGOfcI
| 3,072
| -1
| 0
| null |
|
[
{
"role": "system",
"content": "Determine whether the query describes a specific moment that exists in the given video. If the video contains the described moment, represent the query derived from the original query for video retrieval. If it does not, identify which scene or aspect is missing and which misleading cues caused the mismatch. Then rewrite the query by re-weighting the focus: explicitly emphasize the missing or essential elements and de-emphasize the misleading or less relevant ones. Do NOT invent or hallucinate any new details that are not present in the original query. Finally, represent the (original or revised) query for video retrieval."
},
{
"role": "user",
"content": "<video>\nQuery: \"An aerobics instructor, identifiable by her black tank top and black shorts, leads a dynamic class in a gym setting, with participants following her lead and mirrors reflecting their movements, as exercise equipment is visible in the background.\""
},
{
"role": "assistant",
"content": "This video does not contain the queried moment. The retrieved scene looks similar due to both featuring an instructor leading a fitness class in a gym, but it is missing the specific details of the instructor's black tank top and black shorts, the presence of mirrors reflecting the participants' movements, and the mention of exercise equipment in the background. Query Rewrite: An aerobics instructor in a black tank top and black shorts leads a dynamic class in a gym, with mirrors reflecting the participants' movements and exercise equipment visible in the background.. To find the correct video, search for Latent slots: <LAT0> <LAT1> <LAT2> <LAT3> <LAT4> <LAT5> <LAT6> <LAT7> <LAT_EOS>"
}
] |
[
"/hub_data2/dohwan/data/retrieval/activitynet/videos/v_fUivipdikFU.mp4"
] |
neg
|
1714394113.730561_fig
|
An aerobics instructor, identifiable by her black tank top and black shorts, leads a dynamic class in a gym setting, with participants following her lead and mirrors reflecting their movements, as exercise equipment is visible in the background.
|
v_2WeG8QGOfcI
| 3,072
|
v_fUivipdikFU
| 3
| 1
|
An instructor leads a dynamic fitness class in an indoor gym, with participants dressed in athletic wear and standing on black exercise platforms.
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.