3D Visual Grounding on MoniRefer Overall (val)
80.56Accuracy@0.253DSSDBest
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| 3DSSDBestSensor=LiDAR+RGB, Type=Two-stage2025.12 | 80.56 | 64.09 | |
| Moni3DVGSensor=LiDAR+RGB, Type=Two-stage2025.12 | 61.92 | 52.06 | |
| WildReferSensor=LiDAR+RGB, Type=One-stage, venue=ECCV20242025.12 | 61.06 | 14.16 | |
| InstancereferSensor=RGB-D, Type=Two-stage, training with ground-truth boxes=true, venue=ICCV20212025.12 | 45.78 | 8.09 | |
| CityreferSensor=UAV Photogrammetry, Type=Two-stage, training with ground-truth boxes=true, venue=NeurIPS20232025.12 | 45.16 | 9.18 | |
| CatRandGTSensor=LiDAR+RGB, Type=Two-stage2025.12 | 26.66 | 26.63 | |
| 3DSSDRandSensor=LiDAR+RGB, Type=Two-stage2025.12 | 6.2 | 5.56 | |
| ScanReferSensor=RGB-D, Type=Two-stage, venue=ECCV20202025.12 | 5.64 | 3.32 |