Waymo Open Dataset Challenge Reports
3D Detection
Given one or more lidar range images and the associated camera images, produce a set of 3D upright boxes for the objects in the scene. This is a simplified version of the 3D tracking challenge in that it ignores the temporal component. A baseline open source model is linked below for your reference.
3D Tracking
Given a temporal sequence (prior frames) of lidar and camera data, produce a set of 3D upright boxes and the correspondences between boxes across frames. (Entries that use future frames will be ineligible for awards.) You may use the baseline model or object detections linked below, if you wish.
2D Detection
Given a set of camera images, produce a set of 2D boxes for the objects in the scene.
2D Tracking
Given a temporal sequence (prior frames) of camera images, produce a set of 2D boxes and the correspondences between boxes across frames. (Entries that use future frames will be ineligible for awards.) You may use the baseline object detections linked below, if you wish.
domain adaptation
Similar to the 3D Detection Challenge: given one or more range images and the associated camera images, produce a set of 3D upright boxes for the objects in the scene. However, we provide additional segments from a new location and only a subset have labels.