3D Object Detection Models
CenterPoint
Introduced by Yin et al. in Center-based 3D Object Detection and TrackingCenterPoint 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 Tracking
Papers
| Paper | Code | Results | Date | Stars |
|---|
Tasks
| Task | Papers | Share |
|---|---|---|
| 3D Object Detection | 15 | 16.85% |
| Object Detection | 14 | 15.73% |
| Autonomous Driving | 8 | 8.99% |
| Autonomous Vehicles | 6 | 6.74% |
| Object Tracking | 6 | 6.74% |
| 3D Multi-Object Tracking | 5 | 5.62% |
| Multi-Object Tracking | 5 | 5.62% |
| Semantic Segmentation | 3 | 3.37% |
| Sensor Fusion | 2 | 2.25% |
Usage Over Time
Components
| Component | Type |
|
|---|---|---|
| 🤖 No Components Found | You can add them if they exist; e.g. Mask R-CNN uses RoIAlign |
