Order from Lab 6C3851

Driving rare events

Real dashcam and fleet video of rare road events (cut-ins, hard braking, pedestrians stepping out, debris, near misses, collisions), each clip with camera calibration, vehicle signals on the camera clock, an event label and time, and 3D box labels with a stated label source. Commercial training rights, a legal basis for faces, plates and drivers, and nothing synthetic or taken from a public set.

How answering works

Any number of people fill one order. The lab receives one dataset, in its own columns.

Your agent reads the order
It looks at what you hold and says which columns you can fill, and what is missing.
Mora checks what you send
Every record: no copies, no personal data, not already public, not on Mora from someone else.
The lab pays per record it accepts
The lab named the price. You fill as much of the order as you can.

The columns the lab wants

You do not need every column. Mora says which ones each dataset fills.

Column Type What it must hold

video video (H.264/H.265, original encode, 108 20 s before to 10 s after the event, one file per camera

camera_calibration json Intrinsics, distortion and position on the vehicle, per camera

can_signals time series, 50 Hz or more Speed, steering angle, brake, throttle, yaw rate, turn signals, timestamped on the camera clock

gps_imu time series Coarsened position and inertial readings, synced to the video

event_type label Cut-in, hard brake, pedestrian dart-out, debris, wrong-way vehicle, near miss, collision

event_time number (seconds) When in the clip the event starts and ends

trigger text What flagged the event: hard brake, AEB, driver button

boxes_3d label, per frame at 10 Hz Class, centre, size, yaw, track id, occlusion, in the vehicle frame

label_source text Person or model, lidar or camera only, reviewed or not

vehicle_id id Generated id of the recording vehicle

recorded_at date Day of recording

region text Country and road type

device_model text The camera or dashcam that recorded the clip

lidar point clouds, optional Lidar sweeps when the vehicle has one

Column	Type	What it must hold
`video`	video (H.264/H.265, original encode, 108	20 s before to 10 s after the event, one file per camera
`camera_calibration`	json	Intrinsics, distortion and position on the vehicle, per camera
`can_signals`	time series, 50 Hz or more	Speed, steering angle, brake, throttle, yaw rate, turn signals, timestamped on the camera clock
`gps_imu`	time series	Coarsened position and inertial readings, synced to the video
`event_type`	label	Cut-in, hard brake, pedestrian dart-out, debris, wrong-way vehicle, near miss, collision
`event_time`	number (seconds)	When in the clip the event starts and ends
`trigger`	text	What flagged the event: hard brake, AEB, driver button
`boxes_3d`	label, per frame at 10 Hz	Class, centre, size, yaw, track id, occlusion, in the vehicle frame
`label_source`	text	Person or model, lidar or camera only, reviewed or not
`vehicle_id`	id	Generated id of the recording vehicle
`recorded_at`	date	Day of recording
`region`	text	Country and road type
`device_model`	text	The camera or dashcam that recorded the clip
`lidar`	point clouds, optional	Lidar sweeps when the vehicle has one

How much, in the lab's words

250,000 distinct events (about 2,000 hours of 30-second clips) from at least 5,000 vehicles and 3 countries, 20% at night or in rain, with 3D boxes on at least 20,000 events and no vehicle giving more than 1% of the events

clips
signals
events
calibration
boxes-3d
lidar