Neu3D
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Novel View Synthesis | Neu3D (test) | PSNR33.19 | 18 | |
| Dynamic and Semantic 3DGS Active Training | Neu3D average across five dynamic scenes | SSIM0.9239 | 10 | |
| Time-agnostic querying | Neu3D (test) | mIoU87.41 | 10 | |
| 4D Gaussian instance segmentation | Neu3D | Time (min)0.6 | 6 | |
| Novel-view Panoptic Segmentation | Neu3D sear steak | mAcc (Pixel)95.36 | 5 | |
| Novel-view Panoptic Segmentation | Neu3D flame steak | Pixel Acc95.31 | 5 | |
| Novel-view Panoptic Segmentation | Neu3D flame salmon | mAcc (Pixel)91.31 | 5 | |
| Novel-view Panoptic Segmentation | Neu3D cut roasted beef | Pixel Accuracy (mAcc-pix)95.12 | 5 | |
| Novel-view Panoptic Segmentation | Neu3D cook spinach | mAcc (Pixel)96.63 | 5 | |
| Novel-view Panoptic Segmentation | Neu3D coffee martini | mAcc (Pixel)96.07 | 5 | |
| Dynamic and semantic 3DGS | Neu3D average across five dynamic scenes | SSIM0.9239 | 4 | |
| 4D Gaussian Instance Segmentation | Neu3D (test) | coffee_martini mIoU0.9147 | 3 | |
| Novel View Synthesis | Neu3D | PSNR (Coffee Martini)28.3 | 2 |
Showing 13 of 13 rows