Novel View Synthesis on Phototourism Trevi Fountain R/2 downsampled (test)
24.45PSNRWild-GS
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| Wild-GSImage downsampling=2x2024.06 | 24.45 | 0.8081 | 0.1622 | 0.52 | 227 | |
| Wild-GS w/o GlobalGlobal appearance encoding (Global)=false2024.06 | 24.18 | 0.8045 | 0.1702 | 0.52 | 224 | |
| Wild-GS w/o {c'}Triplane components {c'}=false2024.06 | 24.14 | 0.8036 | 0.1657 | 0.47 | 227 | |
| Wild-GS w/o MaskTransient mask prediction (Mask)=false2024.06 | 23.91 | 0.8063 | 0.159 | 0.53 | 205 | |
| Wild-GS w/o CropTriplane cropping (Crop)=false2024.06 | 23.61 | 0.7915 | 0.1856 | 0.35 | 223 | |
| Wild-GS w/o DepthDepth regularization (Depth)=false2024.06 | 23.2 | 0.7878 | 0.1827 | 0.48 | 218 | |
| Wild-GSTraining iterations=15k, Image downsampling=2x2024.06 | 22.77 | 0.7629 | 0.2229 | 0.23 | 225 | |
| CR-NeRFImage downsampling=2x2024.06 | 21.48 | 0.7117 | 0.2069 | 101 | 0.02 | |
| Ha-NeRFImage downsampling=2x2024.06 | 20.18 | 0.6908 | 0.2225 | 71.6 | 0.05 | |
| NeRF-WImage downsampling=2x2024.06 | 18.97 | 0.6984 | 0.2652 | 60.3 | 0.05 | |
| 3DGSImage downsampling=2x2024.06 | 17.23 | 0.6963 | 0.2815 | 0.12 | 220 |