PandaSet
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| 3D Semantic Segmentation | PandaSet nuScenes source 8 classes (val) | mIoU65.7 | 11 | |
| Urban Scene Reconstruction | PandaSet Extrapolation Hard (test) | FID@4m102.73 | 8 | |
| Urban Scene Reconstruction | PandaSet Extrapolation Moderate (test) | FID@1m59.44 | 8 | |
| Urban Scene Reconstruction | PandaSet Interpolation (test) | PSNR24.93 | 8 | |
| LiDAR Reconstruction | PandaSet | Median L2 Error0.002 | 8 | |
| Camera Reconstruction | PandaSet | PSNR29.76 | 8 | |
| Novel View Synthesis | PandaSet | PSNR27.12 | 8 | |
| View Reconstruction and Interpolation | Pandaset | PSNR28.28 | 8 | |
| Generative View Synthesis | Pandaset | FID (2m)57.1 | 8 | |
| 3D Object Detection | PandaSet (val) | AP BEV80.3 | 8 | |
| Novel-view synthesis | PandaSet 10 scenes | PSNR27 | 7 | |
| Scene Reconstruction | PandaSet 10 scenes | PSNR29.05 | 7 | |
| Surface Reconstruction | Pandaset static scenes | P->M Error (Seq 23)0.45 | 6 | |
| 3D Object Detection | PandaSet | mAP36.6 | 6 | |
| View Extrapolation (Lane Shift) | PandaSet (test) | FID @ 2m62.15 | 6 | |
| View Interpolation | PandaSet (test) | PSNR27.54 | 6 | |
| Sparse View Synthesis | PandaSet Sparse View Split (10% frames for train) | PSNR21.34 | 5 | |
| Novel View Synthesis | PandaSet Out-of-Domain (val) | PSNR26.27 | 5 | |
| View Synthesis | Pandaset Lane Shift | FID (2m)106.5 | 5 | |
| View Synthesis | Pandaset Interpolation | PSNR27.255 | 5 | |
| Scene Relighting | PandaSet | FID85.4 | 5 | |
| Unconditional 3D Generation | PandaSet | FID59.5 | 4 | |
| 360° View Synthesis | PandaSet 360° View (Rotating actors 0° to 360°) | FID100.28 | 4 | |
| Novel Camera Synthesis | PandaSet (Novel Camera Split) | PSNR18.36 | 4 | |
| Novel View Synthesis | PandaSet (50% optimization split) | PSNR26.78 | 4 |
Showing 25 of 41 rows