Coordinates Are NOT Lonely - Codebook Prior Helps Implicit Neural 3D Representations Wen Liu

Implicit neural 3D representation has achieved impressive results in surface or scene reconstruction and novel view synthesis, which typically uses the coordinatebased multi-layer perceptrons (MLPs) to learn a continuous scene representation. However, existing approaches, such as Neural Radiance Field (NeRF) [15], and its variants [16, 26, 29], usually require dense input views (i.e.