TAU-R1: Visual Language Model for Traffic Anomaly Understanding
Paper • 2603.19098 • Published
conversations listlengths 2 2 | video stringclasses 300
values | system_prompt stringclasses 5
values |
|---|---|---|
[
{
"from": "human",
"value": "<video>\nReview the full video for anomalous driving and return only a single token A, B, C, or D that best matches the most severe event."
},
{
"from": "gpt",
"value": "B"
}
] | /data/RangelineSMedicalDr_av_117_22_4.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nReview the full video for anomalous driving and return only a single token A, B, C, or D that best matches the most severe event."
},
{
"from": "gpt",
"value": "B"
}
] | /data/RangelineSMedicalDr_av_224_21_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nWatch the complete video, choose the single best label A, B, C, or D based on the most severe anomaly present, and output only that token."
},
{
"from": "gpt",
"value": "C"
}
] | /data/RangelineSMedicalDr_av_253_18_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nReview the full video for anomalous driving and return only a single token A, B, C, or D that best matches the most severe event."
},
{
"from": "gpt",
"value": "C"
}
] | /data/RangelineSMedicalDr_av_341_19_4.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nWatch the complete video, choose the single best label A, B, C, or D based on the most severe anomaly present, and output only that token."
},
{
"from": "gpt",
"value": "B"
}
] | /data/RangelineS116thSt_av_176_21_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nAnalyze the entire traffic video and output exactly one label token A, B, C, or D for the most severe anomaly, with no extra text."
},
{
"from": "gpt",
"value": "B"
}
] | /data/RangelineS116thSt_av_186_21_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nReview the full video for anomalous driving and return only a single token A, B, C, or D that best matches the most severe event."
},
{
"from": "gpt",
"value": "C"
}
] | /data/RangelineS116thSt_av_198_18_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nReview the full video for anomalous driving and return only a single token A, B, C, or D that best matches the most severe event."
},
{
"from": "gpt",
"value": "B"
}
] | /data/RangelineS116thSt_av_305_14_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nInspect all road users across the whole clip and output exactly one label A, B, C, or D for the most severe and relevant anomaly, nothing else."
},
{
"from": "gpt",
"value": "C"
}
] | /data/RangelineS116thSt_av_53_18_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nReview the full video for anomalous driving and return only a single token A, B, C, or D that best matches the most severe event."
},
{
"from": "gpt",
"value": "C"
}
] | /data/Car-Car-PE-HazelDell-116TH-09-13-24.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nReview the full video for anomalous driving and return only a single token A, B, C, or D that best matches the most severe event."
},
{
"from": "gpt",
"value": "C"
}
] | /data/Car-Car-PE-Keystone-116TH-09-03-25.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nReview the full video for anomalous driving and return only a single token A, B, C, or D that best matches the most severe event."
},
{
"from": "gpt",
"value": "C"
}
] | /data/Car-Car-R3-Keystone-116TH-08-18-25.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nWatch the complete video, choose the single best label A, B, C, or D based on the most severe anomaly present, and output only that token."
},
{
"from": "gpt",
"value": "C"
}
] | /data/Car-Rndbt-DUI-Keystone-106TH-07-19-25.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nDetermine if any anomalies occur in the video and output only one token A, B, C, or D corresponding to the most severe anomaly observed."
},
{
"from": "gpt",
"value": "C"
}
] | /data/Car-Rndbt-Shelborne-126TH-11-04-23-Short.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nAnalyze the entire traffic video and output exactly one label token A, B, C, or D for the most severe anomaly, with no extra text."
},
{
"from": "gpt",
"value": "C"
}
] | /data/Car-Car-NoStop-CityCenter-CarmelDr-12-20-25.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nReview the full video for anomalous driving and return only a single token A, B, C, or D that best matches the most severe event."
},
{
"from": "gpt",
"value": "B"
}
] | /data/no8_cat22_sev4.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nInspect all road users across the whole clip and output exactly one label A, B, C, or D for the most severe and relevant anomaly, nothing else."
},
{
"from": "gpt",
"value": "B"
}
] | /data/no24_cat13_sev4.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nWatch the complete video, choose the single best label A, B, C, or D based on the most severe anomaly present, and output only that token."
},
{
"from": "gpt",
"value": "B"
}
] | /data/no16_cat23_sev4_rm.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nReview the full video for anomalous driving and return only a single token A, B, C, or D that best matches the most severe event."
},
{
"from": "gpt",
"value": "B"
}
] | /data/no16_cat23_sev4_rc.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nReview the full video for anomalous driving and return only a single token A, B, C, or D that best matches the most severe event."
},
{
"from": "gpt",
"value": "D"
}
] | /data/no6_cat5_sev2.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nReview the full video for anomalous driving and return only a single token A, B, C, or D that best matches the most severe event."
},
{
"from": "gpt",
"value": "C"
}
] | /data/no5_cat19_sev4.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nAnalyze the entire traffic video and output exactly one label token A, B, C, or D for the most severe anomaly, with no extra text."
},
{
"from": "gpt",
"value": "C"
}
] | /data/RangelineSMedicalDr_av_116_18_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nReview the full video for anomalous driving and return only a single token A, B, C, or D that best matches the most severe event."
},
{
"from": "gpt",
"value": "C"
}
] | /data/RangelineSMedicalDr_av_127_18_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nWatch the complete video, choose the single best label A, B, C, or D based on the most severe anomaly present, and output only that token."
},
{
"from": "gpt",
"value": "B"
}
] | /data/RangelineSMedicalDr_av_132_21_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nWatch the complete video, choose the single best label A, B, C, or D based on the most severe anomaly present, and output only that token."
},
{
"from": "gpt",
"value": "B"
}
] | /data/RangelineSMedicalDr_av_133_21_4.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nInspect all road users across the whole clip and output exactly one label A, B, C, or D for the most severe and relevant anomaly, nothing else."
},
{
"from": "gpt",
"value": "B"
}
] | /data/RangelineSMedicalDr_av_140_21_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nReview the full video for anomalous driving and return only a single token A, B, C, or D that best matches the most severe event."
},
{
"from": "gpt",
"value": "C"
}
] | /data/RangelineSMedicalDr_av_160_18_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nDetermine if any anomalies occur in the video and output only one token A, B, C, or D corresponding to the most severe anomaly observed."
},
{
"from": "gpt",
"value": "C"
}
] | /data/RangelineSMedicalDr_av_162_30_2.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nReview the full video for anomalous driving and return only a single token A, B, C, or D that best matches the most severe event."
},
{
"from": "gpt",
"value": "D"
}
] | /data/RangelineSMedicalDr_av_181_27_1.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nReview the full video for anomalous driving and return only a single token A, B, C, or D that best matches the most severe event."
},
{
"from": "gpt",
"value": "C"
}
] | /data/RangelineSMedicalDr_av_189_18_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nInspect all road users across the whole clip and output exactly one label A, B, C, or D for the most severe and relevant anomaly, nothing else."
},
{
"from": "gpt",
"value": "D"
}
] | /data/RangelineSMedicalDr_av_195_27_1.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nReview the full video for anomalous driving and return only a single token A, B, C, or D that best matches the most severe event."
},
{
"from": "gpt",
"value": "B"
}
] | /data/RangelineSMedicalDr_av_200_21_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nReview the full video for anomalous driving and return only a single token A, B, C, or D that best matches the most severe event."
},
{
"from": "gpt",
"value": "B"
}
] | /data/RangelineSMedicalDr_av_211_14_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nReview the full video for anomalous driving and return only a single token A, B, C, or D that best matches the most severe event."
},
{
"from": "gpt",
"value": "D"
}
] | /data/RangelineSMedicalDr_av_216_27_1.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nAnalyze the entire traffic video and output exactly one label token A, B, C, or D for the most severe anomaly, with no extra text."
},
{
"from": "gpt",
"value": "C"
}
] | /data/RangelineSMedicalDr_av_218_18_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nWatch the complete video, choose the single best label A, B, C, or D based on the most severe anomaly present, and output only that token."
},
{
"from": "gpt",
"value": "C"
}
] | /data/RangelineSMedicalDr_av_230_18_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nAnalyze the entire traffic video and output exactly one label token A, B, C, or D for the most severe anomaly, with no extra text."
},
{
"from": "gpt",
"value": "C"
}
] | /data/RangelineSMedicalDr_av_236_18_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nAnalyze the entire traffic video and output exactly one label token A, B, C, or D for the most severe anomaly, with no extra text."
},
{
"from": "gpt",
"value": "C"
}
] | /data/RangelineSMedicalDr_av_23_18_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nWatch the complete video, choose the single best label A, B, C, or D based on the most severe anomaly present, and output only that token."
},
{
"from": "gpt",
"value": "D"
}
] | /data/RangelineSMedicalDr_av_240_28_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nWatch the complete video, choose the single best label A, B, C, or D based on the most severe anomaly present, and output only that token."
},
{
"from": "gpt",
"value": "B"
}
] | /data/RangelineSMedicalDr_av_242_21_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nDetermine if any anomalies occur in the video and output only one token A, B, C, or D corresponding to the most severe anomaly observed."
},
{
"from": "gpt",
"value": "B"
}
] | /data/RangelineSMedicalDr_av_250_21_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nReview the full video for anomalous driving and return only a single token A, B, C, or D that best matches the most severe event."
},
{
"from": "gpt",
"value": "C"
}
] | /data/RangelineSMedicalDr_av_252_18_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nReview the full video for anomalous driving and return only a single token A, B, C, or D that best matches the most severe event."
},
{
"from": "gpt",
"value": "C"
}
] | /data/RangelineSMedicalDr_av_253_30_2.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nDetermine if any anomalies occur in the video and output only one token A, B, C, or D corresponding to the most severe anomaly observed."
},
{
"from": "gpt",
"value": "C"
}
] | /data/RangelineSMedicalDr_av_257_19_4.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nInspect all road users across the whole clip and output exactly one label A, B, C, or D for the most severe and relevant anomaly, nothing else."
},
{
"from": "gpt",
"value": "C"
}
] | /data/RangelineSMedicalDr_av_258_18_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nAnalyze the entire traffic video and output exactly one label token A, B, C, or D for the most severe anomaly, with no extra text."
},
{
"from": "gpt",
"value": "B"
}
] | /data/RangelineSMedicalDr_av_263_21_4.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nDetermine if any anomalies occur in the video and output only one token A, B, C, or D corresponding to the most severe anomaly observed."
},
{
"from": "gpt",
"value": "B"
}
] | /data/RangelineSMedicalDr_av_263_22_4.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nWatch the complete video, choose the single best label A, B, C, or D based on the most severe anomaly present, and output only that token."
},
{
"from": "gpt",
"value": "D"
}
] | /data/RangelineSMedicalDr_av_268_27_1.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nReview the full video for anomalous driving and return only a single token A, B, C, or D that best matches the most severe event."
},
{
"from": "gpt",
"value": "C"
}
] | /data/RangelineSMedicalDr_av_273_30_2.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nWatch the complete video, choose the single best label A, B, C, or D based on the most severe anomaly present, and output only that token."
},
{
"from": "gpt",
"value": "C"
}
] | /data/RangelineSMedicalDr_av_278_18_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nDetermine if any anomalies occur in the video and output only one token A, B, C, or D corresponding to the most severe anomaly observed."
},
{
"from": "gpt",
"value": "B"
}
] | /data/RangelineSMedicalDr_av_279_22_4.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nWatch the complete video, choose the single best label A, B, C, or D based on the most severe anomaly present, and output only that token."
},
{
"from": "gpt",
"value": "C"
}
] | /data/RangelineSMedicalDr_av_280_30_2.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nReview the full video for anomalous driving and return only a single token A, B, C, or D that best matches the most severe event."
},
{
"from": "gpt",
"value": "D"
}
] | /data/RangelineSMedicalDr_av_283_27_1.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nDetermine if any anomalies occur in the video and output only one token A, B, C, or D corresponding to the most severe anomaly observed."
},
{
"from": "gpt",
"value": "D"
}
] | /data/RangelineSMedicalDr_av_289_27_1.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nAnalyze the entire traffic video and output exactly one label token A, B, C, or D for the most severe anomaly, with no extra text."
},
{
"from": "gpt",
"value": "D"
}
] | /data/RangelineSMedicalDr_av_28_5_2.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nWatch the complete video, choose the single best label A, B, C, or D based on the most severe anomaly present, and output only that token."
},
{
"from": "gpt",
"value": "C"
}
] | /data/RangelineSMedicalDr_av_293_18_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nInspect all road users across the whole clip and output exactly one label A, B, C, or D for the most severe and relevant anomaly, nothing else."
},
{
"from": "gpt",
"value": "C"
}
] | /data/RangelineSMedicalDr_av_29_18_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nWatch the complete video, choose the single best label A, B, C, or D based on the most severe anomaly present, and output only that token."
},
{
"from": "gpt",
"value": "C"
}
] | /data/RangelineSMedicalDr_av_312_18_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nWatch the complete video, choose the single best label A, B, C, or D based on the most severe anomaly present, and output only that token."
},
{
"from": "gpt",
"value": "C"
}
] | /data/RangelineSMedicalDr_av_31_18_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nWatch the complete video, choose the single best label A, B, C, or D based on the most severe anomaly present, and output only that token."
},
{
"from": "gpt",
"value": "B"
}
] | /data/RangelineSMedicalDr_av_338_21_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nAnalyze the entire traffic video and output exactly one label token A, B, C, or D for the most severe anomaly, with no extra text."
},
{
"from": "gpt",
"value": "D"
}
] | /data/RangelineSMedicalDr_av_35_5_2.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nAnalyze the entire traffic video and output exactly one label token A, B, C, or D for the most severe anomaly, with no extra text."
},
{
"from": "gpt",
"value": "C"
}
] | /data/RangelineSMedicalDr_av_37_19_4.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nAnalyze the entire traffic video and output exactly one label token A, B, C, or D for the most severe anomaly, with no extra text."
},
{
"from": "gpt",
"value": "C"
}
] | /data/RangelineSMedicalDr_av_38_18_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nDetermine if any anomalies occur in the video and output only one token A, B, C, or D corresponding to the most severe anomaly observed."
},
{
"from": "gpt",
"value": "C"
}
] | /data/RangelineSMedicalDr_av_44_18_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nReview the full video for anomalous driving and return only a single token A, B, C, or D that best matches the most severe event."
},
{
"from": "gpt",
"value": "D"
}
] | /data/RangelineSMedicalDr_av_65_27_1.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nDetermine if any anomalies occur in the video and output only one token A, B, C, or D corresponding to the most severe anomaly observed."
},
{
"from": "gpt",
"value": "C"
}
] | /data/RangelineSMedicalDr_av_80_18_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nInspect all road users across the whole clip and output exactly one label A, B, C, or D for the most severe and relevant anomaly, nothing else."
},
{
"from": "gpt",
"value": "D"
}
] | /data/RangelineSMedicalDr_av_93_27_1.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nWatch the complete video, choose the single best label A, B, C, or D based on the most severe anomaly present, and output only that token."
},
{
"from": "gpt",
"value": "D"
}
] | /data/RangelineSMedicalDr_av_98_27_1.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nReview the full video for anomalous driving and return only a single token A, B, C, or D that best matches the most severe event."
},
{
"from": "gpt",
"value": "C"
}
] | /data/RangelineS116thSt_av_111_19_4.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nDetermine if any anomalies occur in the video and output only one token A, B, C, or D corresponding to the most severe anomaly observed."
},
{
"from": "gpt",
"value": "B"
}
] | /data/RangelineS116thSt_av_114_21_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nDetermine if any anomalies occur in the video and output only one token A, B, C, or D corresponding to the most severe anomaly observed."
},
{
"from": "gpt",
"value": "B"
}
] | /data/RangelineS116thSt_av_116_21_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nInspect all road users across the whole clip and output exactly one label A, B, C, or D for the most severe and relevant anomaly, nothing else."
},
{
"from": "gpt",
"value": "B"
}
] | /data/RangelineS116thSt_av_121_21_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nReview the full video for anomalous driving and return only a single token A, B, C, or D that best matches the most severe event."
},
{
"from": "gpt",
"value": "C"
}
] | /data/RangelineS116thSt_av_12_18_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nDetermine if any anomalies occur in the video and output only one token A, B, C, or D corresponding to the most severe anomaly observed."
},
{
"from": "gpt",
"value": "B"
}
] | /data/RangelineS116thSt_av_142_21_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nAnalyze the entire traffic video and output exactly one label token A, B, C, or D for the most severe anomaly, with no extra text."
},
{
"from": "gpt",
"value": "C"
}
] | /data/RangelineS116thSt_av_153_18_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nWatch the complete video, choose the single best label A, B, C, or D based on the most severe anomaly present, and output only that token."
},
{
"from": "gpt",
"value": "C"
}
] | /data/RangelineS116thSt_av_169_18_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nInspect all road users across the whole clip and output exactly one label A, B, C, or D for the most severe and relevant anomaly, nothing else."
},
{
"from": "gpt",
"value": "C"
}
] | /data/RangelineS116thSt_av_16_18_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nWatch the complete video, choose the single best label A, B, C, or D based on the most severe anomaly present, and output only that token."
},
{
"from": "gpt",
"value": "C"
}
] | /data/RangelineS116thSt_av_17_18_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nDetermine if any anomalies occur in the video and output only one token A, B, C, or D corresponding to the most severe anomaly observed."
},
{
"from": "gpt",
"value": "C"
}
] | /data/RangelineS116thSt_av_183_18_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nDetermine if any anomalies occur in the video and output only one token A, B, C, or D corresponding to the most severe anomaly observed."
},
{
"from": "gpt",
"value": "B"
}
] | /data/RangelineS116thSt_av_188_14_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nWatch the complete video, choose the single best label A, B, C, or D based on the most severe anomaly present, and output only that token."
},
{
"from": "gpt",
"value": "C"
}
] | /data/RangelineS116thSt_av_189_32_4.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nInspect all road users across the whole clip and output exactly one label A, B, C, or D for the most severe and relevant anomaly, nothing else."
},
{
"from": "gpt",
"value": "D"
}
] | /data/RangelineS116thSt_av_201_5_2.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nInspect all road users across the whole clip and output exactly one label A, B, C, or D for the most severe and relevant anomaly, nothing else."
},
{
"from": "gpt",
"value": "C"
}
] | /data/RangelineS116thSt_av_222_18_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nInspect all road users across the whole clip and output exactly one label A, B, C, or D for the most severe and relevant anomaly, nothing else."
},
{
"from": "gpt",
"value": "C"
}
] | /data/RangelineS116thSt_av_233_18_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nWatch the complete video, choose the single best label A, B, C, or D based on the most severe anomaly present, and output only that token."
},
{
"from": "gpt",
"value": "B"
}
] | /data/RangelineS116thSt_av_235_21_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nDetermine if any anomalies occur in the video and output only one token A, B, C, or D corresponding to the most severe anomaly observed."
},
{
"from": "gpt",
"value": "B"
}
] | /data/RangelineS116thSt_av_237_21_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nWatch the complete video, choose the single best label A, B, C, or D based on the most severe anomaly present, and output only that token."
},
{
"from": "gpt",
"value": "B"
}
] | /data/RangelineS116thSt_av_244_21_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nInspect all road users across the whole clip and output exactly one label A, B, C, or D for the most severe and relevant anomaly, nothing else."
},
{
"from": "gpt",
"value": "B"
}
] | /data/RangelineS116thSt_av_253_21_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nReview the full video for anomalous driving and return only a single token A, B, C, or D that best matches the most severe event."
},
{
"from": "gpt",
"value": "C"
}
] | /data/RangelineS116thSt_av_270_30_2.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nWatch the complete video, choose the single best label A, B, C, or D based on the most severe anomaly present, and output only that token."
},
{
"from": "gpt",
"value": "B"
}
] | /data/RangelineS116thSt_av_276_23_4.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nAnalyze the entire traffic video and output exactly one label token A, B, C, or D for the most severe anomaly, with no extra text."
},
{
"from": "gpt",
"value": "C"
}
] | /data/RangelineS116thSt_av_277_18_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nInspect all road users across the whole clip and output exactly one label A, B, C, or D for the most severe and relevant anomaly, nothing else."
},
{
"from": "gpt",
"value": "C"
}
] | /data/RangelineS116thSt_av_27_18_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nWatch the complete video, choose the single best label A, B, C, or D based on the most severe anomaly present, and output only that token."
},
{
"from": "gpt",
"value": "B"
}
] | /data/RangelineS116thSt_av_280_21_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nWatch the complete video, choose the single best label A, B, C, or D based on the most severe anomaly present, and output only that token."
},
{
"from": "gpt",
"value": "C"
}
] | /data/RangelineS116thSt_av_281_18_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nReview the full video for anomalous driving and return only a single token A, B, C, or D that best matches the most severe event."
},
{
"from": "gpt",
"value": "C"
}
] | /data/RangelineS116thSt_av_286_18_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nWatch the complete video, choose the single best label A, B, C, or D based on the most severe anomaly present, and output only that token."
},
{
"from": "gpt",
"value": "B"
}
] | /data/RangelineS116thSt_av_287_21_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nInspect all road users across the whole clip and output exactly one label A, B, C, or D for the most severe and relevant anomaly, nothing else."
},
{
"from": "gpt",
"value": "C"
}
] | /data/RangelineS116thSt_av_289_18_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nDetermine if any anomalies occur in the video and output only one token A, B, C, or D corresponding to the most severe anomaly observed."
},
{
"from": "gpt",
"value": "B"
}
] | /data/RangelineS116thSt_av_294_21_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nWatch the complete video, choose the single best label A, B, C, or D based on the most severe anomaly present, and output only that token."
},
{
"from": "gpt",
"value": "B"
}
] | /data/RangelineS116thSt_av_29_21_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
[
{
"from": "human",
"value": "<video>\nAnalyze the entire traffic video and output exactly one label token A, B, C, or D for the most severe anomaly, with no extra text."
},
{
"from": "gpt",
"value": "C"
}
] | /data/RangelineS116thSt_av_308_18_3.mp4 |
You are a video classification system that identifies anomalous driving behavior in traffic videos.
You must assign exactly one label token from the set {A, B, C, D} based on all visible road users across the entire video. Choose the single best label using the definitions below. If multiple anomalies occur, output only the most severe and more relevant one.
Labels (output exactly one token):
- A: No anomaly. All visible road users behave normally.
- B: Direction or maneuver violation, including wrong-way driving, illegal or unsafe turns, and illegal or unsafe lane changes.
- C: Near-collision or collision event, including cut-offs that force hard braking or evasive action, near-misses, and crashes.
- D: Abnormal road use, including stopping in an active lane or mid-road (breakdown, yielding to emergency vehicles, letting pedestrians cross) and driving or parking in non-vehicle areas (sidewalk, shoulder, etc.).
When you answer, you MUST:
1. Watch the entire video before deciding.
2. Output exactly one label token only: A or B or C or D.
3. Output nothing else: no words, no punctuation, no extra whitespace, no newlines.
|
End of preview. Expand in Data Studio
Roundabout-TAU: A Real-World Traffic Anomaly Understanding Dataset
Overview
Roundabout-TAU is a real-world traffic anomaly dataset that support multiple tasks and mainly support Multi-Modality Large Language Models(MLLMs) traffic anomaly understanding. We collect 342 video clips, includes 276 anomaly clips and 66 normal clips, from 28 fixed roadside cameras through the cooperation with the City of Carmel, Indiana. We further provide test-based QA annotations to support training for MLLMs.
Dataset Contents
/videos.zip: contains all 342 videos./annotations/roundabout_vau_summary_qa_pairs.json: Anomaly summarization task for all anomalous videos. This requires the model to summarize the anomaly event from the perspectives of environment, object grounding, description, and reasoning, forming a coherent paragraph./annotations/roundabout_vau_category_qa_pairs.json: Anomaly classification/detection task for all videos. The dataset provides 4 anomaly category labels: A: No anomaly, B: Direction or maneuver violation, C: Near-collision or collision, D: Abnormal road use. Further definitions of these categories are provided in the system prompt./annotations/roundabout_vau_object_grounding_qa_pairs.json: Anomaly object grounding task for anomalous videos. This annotation includes environmental and frame-level grounding./annotations/roundabout_vau_time_window_qa_pairs.json: Anomaly temporal grounding task for anomalous videos. The task requires the model to generate a temporal window [start_sec, end_sec] indicating the duration of the anomaly event./annotations/roundabout_vau_simple_description_qa_pairs.json: Anomaly description task for anomalous videos. This requires the model to summarize the anomaly event in one simple sentence./annotations/roundabout_vau_simple_reasoning_qa_pairs.json: Anomaly reasoning task for anomalous videos. This requires the model to provide a simple explanation of why the anomaly event occurred./annotations/roundabout_vau_environment_qa_pairs.json: Environmental perception task for all videos. This asks the model to identify road type, weather, road surface, and time of day./annotations/roundabout_vau_all_qa_pairs.json: A collection of all annotated QA pairs.
Dataset Split Information
We randomly split dataset into training (300 videos) and test sets (42 videos) to test our TAU-R1 framework. The splits are documented in /annotations/roundabout_tau_train_videos.txt and /annotations/roundabout_tau_test_videos.txt. The corresponding JSON annotations for the train/test splits are in /splited_annotations.
How to Use
- download/extract all videos sources in corresponding folder.
- download task-specific json and change video path to your corresponding video directory.
More Details(Paper)
More details of this dataset can be got from paper TAU-R1.
License
This dataset is released under the MIT License. You are free to use, modify, and distribute this dataset in accordance with the terms of the MIT License.
See the LICENSE file for more details.
Citation
If you use this dataset in your research, please cite the following paper:
@misc{lin2026taur1visuallanguagemodel,
title = {TAU-R1: Visual Language Model for Traffic Anomaly Understanding},
author = {Yuqiang Lin and Kehua Chen and Sam Lockyer and Arjun Yadav and Mingxuan Sui and Shucheng Zhang and Yan Shi and Bingzhang Wang and Yuang Zhang and Markus Zarbock and Florain Stanek and Adrian Evans and Wenbin Li and Yinhai Wang and Nic Zhang},
year = {2026},
eprint = {2603.19098},
archivePrefix= {arXiv},
primaryClass = {cs.CV},
url = {https://arxiv.org/abs/2603.19098}
}
- Downloads last month
- 42