3D Video Reconstruction on Youtube video dataset 64x64x64 grid, 12.5% random sampling (test)
26.44PSNRTri
Evaluation results
| Method | Links | |
|---|---|---|
| TriNo. of params (memory)=1, 445, 891 (5.78M), Time (s)=352.66, Network Architecture=Shifted triangle encoding (simple) with Deep (4 layer) MLPs2022.05 | 26.44 | |
| LinFNo. of params (memory)=1, 445, 891 (5.78M), Time (s)=349.49, Network Architecture=Simple positional encoding with Deep (4 layer) MLPs2022.05 | 26.15 | |
| LogFNo. of params (memory)=1, 445, 891 (5.78M), Time (s)=350.62, Network Architecture=Simple positional encoding with Deep (4 layer) MLPs2022.05 | 25.63 | |
| RFFNo. of params (memory)=1, 445, 891 (5.78M), Time (s)=350.92, Network Architecture=Simple positional encoding with Deep (4 layer) MLPs2022.05 | 25.46 | |
| GauNo. of params (memory)=1, 445, 891 (5.78M), Time (s)=351.45, Network Architecture=Shifted Gaussian encoding (simple) with Deep (4 layer) MLPs2022.05 | 24.58 | |
| GauNo. of params (memory)=786, 432 (3.15M), Time (s)=3.46, Network Architecture=Complex shifted-based encoding (Gau) with single linear layer and smart indexing2022.05 | 24.44 | |
| TriNo. of params (memory)=786, 432 (3.15M), Time (s)=3.39, Network Architecture=Complex shifted-based encoding (Tri) with single linear layer and smart indexing2022.05 | 23.91 | |
| RFFNo. of params (memory)=786, 432 (3.15M), Time (s)=3.62, Network Architecture=Complex positional encoding with single linear layer and smart indexing2022.05 | 21.4 | |
| LogFNo. of params (memory)=786, 432 (3.15M), Time (s)=3.41, Network Architecture=Complex positional encoding with single linear layer and smart indexing2022.05 | 16.76 | |
| LinFNo. of params (memory)=786, 432 (3.15M), Time (s)=3.37, Network Architecture=Complex positional encoding with single linear layer and smart indexing2022.05 | 12.87 |