CenterPoint is a two-stage 3D detector that finds centers of objects and their properties using a keypoint detector and regresses to other attributes, including 3D size, 3D orientation and velocity. In a second-stage, it refines these estimates using additional point features on the object. CenterPoint uses a standard Lidar-based backbone network, i.e., VoxelNet or PointPillars, to build a representation of the input point-cloud. CenterPoint predicts the relative offset (velocity) of objects between consecutive frames, which are then linked up greedily -- so in Centerpoint, 3D object tracking simplifies to greedy closest-point matching.
Source: Center-based 3D Object Detection and TrackingPaper | Code | Results | Date | Stars |
---|
Task | Papers | Share |
---|---|---|
3D Object Detection | 15 | 16.48% |
Object Detection | 14 | 15.38% |
Autonomous Driving | 8 | 8.79% |
Autonomous Vehicles | 6 | 6.59% |
Object Tracking | 6 | 6.59% |
3D Multi-Object Tracking | 5 | 5.49% |
Multi-Object Tracking | 5 | 5.49% |
Semantic Segmentation | 3 | 3.30% |
Sensor Fusion | 2 | 2.20% |
Component | Type |
|
---|---|---|
🤖 No Components Found | You can add them if they exist; e.g. Mask R-CNN uses RoIAlign |