Novel View Synthesis on TanksAndTemples Low Data (PSNR, SSIM, LPIPS, AVG-E)
20.578PSNROurs
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| OursBackbone=3DGS2026.04 | 20.578 | 0.7399 | 0.273 | 0.1083 | |
| OursBackbone=2DGS2026.04 | 20.35 | 0.7328 | 0.2933 | 0.1136 | |
| FSGSBackbone=3DGS2026.04 | 20.307 | 0.7213 | 0.3069 | 0.1165 | |
| LsidBackbone=3DGS2026.04 | 20.173 | 0.7208 | 0.2721 | 0.1132 | |
| SparseGSBackbone=3DGS2026.04 | 20.152 | 0.7202 | 0.2719 | 0.1134 | |
| LsidBackbone=2DGS2026.04 | 20.075 | 0.7259 | 0.2787 | 0.1148 | |
| 2DGS (w/o depth)Backbone=2DGS2026.04 | 19.931 | 0.7261 | 0.2788 | 0.1163 | |
| 3DGS (w/o depth)Backbone=3DGS2026.04 | 19.916 | 0.7206 | 0.2715 | 0.1156 | |
| DNGaussianBackbone=3DGS2026.04 | 19.74 | 0.6905 | 0.3717 | 0.1314 |