Object Detection on ScanNet 20 (val)
43.3mAP@50UniPre3D
Evaluation results
| Method | Links | ||
|---|---|---|---|
| UniPre3DBackbone=VoteNet, Pre-train=UniPre3D2025.06 | 43.3 | 64 | |
| Ponder-RGBDBackbone=VoteNet, Pre-train=Ponder-RGBD2025.06 | 41 | 63.6 | |
| IAEBackbone=VoteNet, Pre-train=IAE2025.06 | 39.8 | 61.5 | |
| DepthContrastBackbone=VoteNet, Pre-train=DepthContrast2025.06 | 39.1 | 62.1 | |
| STRLBackbone=VoteNet, Pre-train=STRL2025.06 | 38.4 | 59.5 | |
| PC-FractalDBBackbone=VoteNet, Pre-train=PC-FractalDB2025.06 | 38.3 | 61.9 | |
| PointContrastBackbone=VoteNet, Pre-train=PointContrast2025.06 | 38 | 59.2 | |
| RandomRoomsBackbone=VoteNet, Pre-train=RandomRooms2025.06 | 36.2 | 61.3 | |
| VoteNet (Scratch)Backbone=VoteNet, Pre-train=None2025.06 | 33.5 | 58.6 |