Across the three displayed Lego viewpoints, the positional and Gaussian mappings produce sharper reconstructed scene details than the no-mapping and basic tiny NeRF models. · CiteArk
Across the three displayed Lego viewpoints, the positional and Gaussian mappings produce sharper reconstructed scene details than the no-mapping and basic tiny NeRF models.
来源：paper:Figure 19 p. 24; Figure 1(e) p. 2; §E.5 pp. 20–21
