Eye on AI Weekly Research Watch
Less Decoder is More Encoder: Geometric Representation Learning from Novel View Synthesis
· 3 min · Season 1
Show notes
Novel view synthesis should teach models about 3D structure, yet encoderbased approaches produce weak representations. The authors blame two architectural choices: decoders that are too spatially expressive, which dilute the encoder's role, and pixel-space targets, which hinder feature learning. SNAP, a self-supervised transformer, uses a pose-conditioned local decoder and a latent-space reconstruction objective. It competes with geometry-supervised methods across localization, pose estimation, correspondence, depth estimation, and robot manipulation. Its features show emergent viewpoint…
Website: share.transistor.fm
