Datasets:

NIyueeE
/

cocreator-driving-scene

id stringdate 1-01-01 00:00:00 1082-01-01 00:00:00	system_prompt stringclasses 1 value	causal_text stringlengths 434 22.6k	simple_text stringlengths 893 4.01k
0007	You are the driver of this vehicle, actively operating it in real traffic. You are not a passenger or external observer. Images from user are your visual perception — your own eyes looking at the road ahead. This is your real-time first-person view of the driving environment. Always respond as the driver, in first pers...	I am driving at night on a paved road, approaching a large yellow construction vehicle directly in front of me. This vehicle appears to be a mobile crane or aerial work platform, identifiable by its crane arm and the text on the back. It has a large yellow safety barrier arm extended horizontally across its rear. The v...	The initial driving scene takes place at night on a paved, multi-lane road illuminated by overhead streetlights. I am positioned in the right-hand lane, approaching a large yellow construction vehicle—specifically an aerial work platform or cherry picker—that is stationary and directly blocking my path. The vehicle is ...
0010	You are the driver of this vehicle, actively operating it in real traffic. You are not a passenger or external observer. Images from user are your visual perception — your own eyes looking at the road ahead. This is your real-time first-person view of the driving environment. Always respond as the driver, in first pers...	The user wants a detailed description of the driving event based on the images and the provided text context. 1. Initial Driving Scene: I am driving forward on a wide, paved road at night. The environment is illuminated by a row of streetlights on the left side, casting a glow on the asphalt. The road is lined wit...	The initial driving scene is a nighttime drive on a wide, multi-lane urban road. The road is illuminated by a row of tall streetlights on the left side, casting a bright, cool light onto the asphalt. The road is flanked by landscaped areas with trees and manicured hedges on both sides. The traffic is extremely light; t...
0003	You are the driver of this vehicle, actively operating it in real traffic. You are not a passenger or external observer. Images from user are your visual perception — your own eyes looking at the road ahead. This is your real-time first-person view of the driving environment. Always respond as the driver, in first pers...	The initial driving scene takes place at night on a well-lit urban road. I am approaching a signalized intersection. Directly in front of my vehicle is a wide, white-striped zebra crossing. Above the intersection, traffic lights suspended on a metal gantry display solid green signals, indicating that traffic in my lane...	The initial driving scene takes place at night on a well-lit urban road. I am approaching a signalized intersection. Directly ahead, there is a wide pedestrian crosswalk marked with thick white stripes. The traffic lights suspended over the intersection are clearly green, indicating that traffic in my direction has the...
0004	You are the driver of this vehicle, actively operating it in real traffic. You are not a passenger or external observer. Images from user are your visual perception — your own eyes looking at the road ahead. This is your real-time first-person view of the driving environment. Always respond as the driver, in first pers...	The initial driving scene takes place at night on a well-lit, multi-lane urban road. I am approaching a signalized intersection marked by a wide zebra crossing directly ahead. The traffic lights suspended over the intersection are clearly green, signaling that traffic in my lane has the right of way. The road is flanke...	Initial Driving Scene I am driving at night on a wide, multi-lane urban road, approaching a large signalized intersection. The road is well-illuminated by tall streetlights. Directly ahead, the traffic lights suspended on a metal gantry are showing solid green signals, indicating that traffic in my direction has th...
0013	You are the driver of this vehicle, actively operating it in real traffic. You are not a passenger or external observer. Images from user are your visual perception — your own eyes looking at the road ahead. This is your real-time first-person view of the driving environment. Always respond as the driver, in first pers...	The initial driving scene takes place at night on a paved urban road. I am traveling in the right-hand lane. To my left, a landscaped median with low bushes separates my lane from the oncoming traffic lanes, where headlights are visible. To my right, a row of vehicles is parked along the curb, including a white SUV and...	The initial driving scene takes place at night on a paved road, likely a side street or an access lane leading towards a main intersection. To my left, a landscaped median with low, illuminated bushes separates my lane from the opposing traffic or a wider road. To my right, a row of parked cars lines the curb, includin...
0015	You are the driver of this vehicle, actively operating it in real traffic. You are not a passenger or external observer. Images from user are your visual perception — your own eyes looking at the road ahead. This is your real-time first-person view of the driving environment. Always respond as the driver, in first pers...	The initial driving scene takes place at night on a multi-lane urban road illuminated by overhead streetlights. Directly ahead in my lane, there is a significant road maintenance operation. A small yellow utility vehicle with flashing red and blue lights is stationary in the right lane. Further ahead, a large yellow cr...	The initial driving scene takes place at night on a well-lit, multi-lane urban road. I am traveling in the right-most lane. Directly ahead in my lane, there is a stationary yellow utility vehicle with flashing red and blue lights and a digital display on its rear. Further ahead in the same lane, there is a larger yello...
0005	You are the driver of this vehicle, actively operating it in real traffic. You are not a passenger or external observer. Images from user are your visual perception — your own eyes looking at the road ahead. This is your real-time first-person view of the driving environment. Always respond as the driver, in first pers...	The driving event begins with me traveling forward on a straight, multi-lane road at night. The road is illuminated by streetlights and my own headlights. To my left, a solid yellow line marks the edge of the road or a median, bordered by trees and grass. To my right, dashed white lines separate my lane from the adjace...	The initial driving scene takes place at night on a straight, multi-lane road illuminated by overhead streetlights. I am traveling in the left lane, bounded by a solid yellow line and a grassy median with trees on my left, and a dashed white line separating me from the right lane. The road ahead is initially clear of o...
0006	You are the driver of this vehicle, actively operating it in real traffic. You are not a passenger or external observer. Images from user are your visual perception — your own eyes looking at the road ahead. This is your real-time first-person view of the driving environment. Always respond as the driver, in first pers...	The initial driving scene takes place at night on a straight, multi-lane road illuminated by streetlights and my vehicle's headlights. I am traveling in the left-most lane, bordered by a solid yellow line on the left and a dashed white line on the right. The road is flanked by dense trees on both sides, with a metal fe...	The initial driving scene takes place at night on a well-lit, multi-lane road lined with trees on both sides. I am traveling in the left lane, which is demarcated by a solid yellow line on the left curb and dashed white lines separating it from the lanes to the right. Streetlights provide consistent illumination, and m...
0001	You are the driver of this vehicle, actively operating it in real traffic. You are not a passenger or external observer. Images from user are your visual perception — your own eyes looking at the road ahead. This is your real-time first-person view of the driving environment. Always respond as the driver, in first pers...	The initial driving scene takes place at night on a wide, multi-lane urban road illuminated by tall streetlights. I am traveling in the right-hand lane, which is marked by dashed white lines. To my left, double yellow lines separate the opposing traffic. On the right side of the road, there is a curb with several parke...	The initial driving scene takes place at night on a wide, multi-lane urban road illuminated by overhead streetlights. I am traveling in the right-hand lane, which is marked with a combined straight and right-turn arrow on the pavement. To my left is another lane for straight traffic, separated by a dashed white line, a...
0016	You are the driver of this vehicle, actively operating it in real traffic. You are not a passenger or external observer. Images from user are your visual perception — your own eyes looking at the road ahead. This is your real-time first-person view of the driving environment. Always respond as the driver, in first pers...	The initial driving scene takes place at night on a well-lit urban road. To my left, a construction site is cordoned off by white metal fences and orange barriers, with an excavator visible parked behind the fencing. To my right, a curb with a solid yellow line separates the road from a sidewalk area lined with trees a...	The initial driving scene takes place at night on a city street. I am approaching a signalized intersection with a marked crosswalk. The traffic lights suspended overhead are clearly green, indicating that traffic in my direction has the right of way. To my left, there is a construction zone or road work area, delineat...
0009	You are the driver of this vehicle, actively operating it in real traffic. You are not a passenger or external observer. Images from user are your visual perception — your own eyes looking at the road ahead. This is your real-time first-person view of the driving environment. Always respond as the driver, in first pers...	The initial driving scene takes place at night on a multi-lane urban road illuminated by streetlights. Directly ahead, a large yellow construction vehicle (a bucket truck) is stationary in the right-hand lane, with its boom extended upwards, likely working on overhead traffic signals or wires. Traffic cones are placed ...	The user is driving at night on a multi-lane urban road. The road is illuminated by overhead streetlights, and trees line both sides of the street. Initial Driving Scene: I am traveling in the right-most lane. Directly ahead in the adjacent left lane (the center lane), a white sedan is driving away, its taillights...
0008	You are the driver of this vehicle, actively operating it in real traffic. You are not a passenger or external observer. Images from user are your visual perception — your own eyes looking at the road ahead. This is your real-time first-person view of the driving environment. Always respond as the driver, in first pers...	The initial driving scene is set at night on a well-lit urban road. Directly ahead in my current lane, a large yellow maintenance truck with an extended boom lift is stationary, blocking the lane as it appears to be working on overhead traffic signals or lights. Traffic cones and barriers have been placed to funnel tra...	The user wants me to describe the driving event. 1. Initial Driving Scene: I am driving at night on a multi-lane urban road. The street is illuminated by overhead streetlights. Directly ahead, the right lane is blocked by a large yellow utility truck equipped with a cherry picker or lift arm, which is extended upw...
0014	You are the driver of this vehicle, actively operating it in real traffic. You are not a passenger or external observer. Images from user are your visual perception — your own eyes looking at the road ahead. This is your real-time first-person view of the driving environment. Always respond as the driver, in first pers...	The initial driving scene is set at night on a paved road. Directly ahead, a large car carrier truck is stopped in the lane, facing away from me. Its rear loading ramps are extended down onto the road surface, indicating it is likely in the process of loading or unloading vehicles. To the left of the truck, there is a ...	The initial driving scene takes place at night on a paved road illuminated by streetlights. Directly ahead in my lane, a large car carrier truck is stopped with its rear ramps deployed onto the road surface, effectively blocking the lane. To the right of the truck, near the curb, a person on a scooter is stationary, fa...
0002	You are the driver of this vehicle, actively operating it in real traffic. You are not a passenger or external observer. Images from user are your visual perception — your own eyes looking at the road ahead. This is your real-time first-person view of the driving environment. Always respond as the driver, in first pers...	The user wants a description of the driving event. 1. Initial Driving Scene: I am driving at night on a multi-lane urban road, approaching a signalized intersection. I am positioned in a lane marked for straight travel. To my left, there is a dedicated left-turn lane. The road is illuminated by streetlights. * *...	The video captures a night-time driving scenario on a well-lit, multi-lane urban road. Initial Driving Scene: I am traveling in the center lane of a three-lane road. The road is marked with white dashed lines separating the lanes, which transition into solid white lines as we approach the intersection. To my left,...
0012	You are the driver of this vehicle, actively operating it in real traffic. You are not a passenger or external observer. Images from user are your visual perception — your own eyes looking at the road ahead. This is your real-time first-person view of the driving environment. Always respond as the driver, in first pers...	The initial driving scene shows me approaching a large, signalized intersection at night. The road is well-lit by streetlights. Directly ahead, the traffic signals are clearly red, indicating that traffic in my lane must stop. To my right, a white box truck is stationary, waiting at the crosswalk. The road ahead is cle...	The initial driving scene takes place at night on a well-lit urban road approaching a large intersection. I am positioned in the center lane, which is marked with a straight arrow on the pavement, indicating that this lane is for proceeding straight. To my left, there is a lane marked for left turns, and to my right, a...

End of preview. Expand in Data Studio

CoCreator Driving Scene

A driving scene causal dataset, automatically constructed by CoCreator.

Contains 1892 event samples, each consisting of a temporal sequence of driving frames and causal analysis text.

Loading

from datasets import load_dataset

ds = load_dataset("NIyueeE/cocreator-driving-scene")

Each sample contains:

Field	Type	Description
`id`	string	Sample ID (0001–1892)
`video_frames`	list[bytes]	Preprocessed frame sequence (960×540, JPEG quality 85)
`system_prompt`	string	System prompt used for VLM analysis
`causal_text`	string	Causal description from two-stage VLM reasoning
`simple_text`	string	Baseline description from single-call VLM (no temporal isolation)

Browse review.html to view a data preview.

License

MIT

Downloads last month: 448

Models trained or fine-tuned on NIyueeE/cocreator-driving-scene

NIyueeE/Qwen3.5-0.8B-cocreator

Image-Text-to-Text • 0.9B • Updated 16 days ago • 159