Nexels: Neurally-Textured Surfels for Real-Time Novel View Synthesis with Sparse Geometries
Auto-discovered
Visit resource ↗Description
Nexels: Neurally-Textured Surfels for Real-Time Novel View Synthesis with Sparse Geometries presents a research project page by Victor Rong, Jan Held, Victor Chu, Daniel Rebain, Marc Van Droogenbroeck, Kiriakos N. Kutulakos, Andrea Tagliasacchi, and David B. Lindell on compact novel-view synthesis that decouples geometry and appearance.
It uses surfels for geometry plus a global Instant-NGP neural field and per-primitive colors for appearance, texturing only the K highest-weight primitives per pixel in a two-pass render. The page compares against 3D Gaussian splatting, reports large cuts in primitive count and memory, and includes a custom close-up texture dataset with perceptual metrics.
It is for advanced researchers and engineers in neural rendering or Gaussian splatting who already know splat-based novel view pipelines.
The URL is a free project page with abstract, interactive comparisons, method diagrams, and result figures.
It is useful when you need a sparse, neurally textured surfel alternative to dense Gaussian clouds for high-frequency appearance. Link last verified accessible on 2026-09-29. Link last verified accessible on 2026-09-28.