English

Differentiable Surface Splatting for Point-based Geometry Processing

Graphics 2019-09-05 v3 Neural and Evolutionary Computing

Abstract

We propose Differentiable Surface Splatting (DSS), a high-fidelity differentiable renderer for point clouds. Gradients for point locations and normals are carefully designed to handle discontinuities of the rendering function. Regularization terms are introduced to ensure uniform distribution of the points on the underlying surface. We demonstrate applications of DSS to inverse rendering for geometry synthesis and denoising, where large scale topological changes, as well as small scale detail modifications, are accurately and robustly handled without requiring explicit connectivity, outperforming state-of-the-art techniques. The data and code are at https://github.com/yifita/DSS.

Keywords

Cite

@article{arxiv.1906.04173,
  title  = {Differentiable Surface Splatting for Point-based Geometry Processing},
  author = {Wang Yifan and Felice Serena and Shihao Wu and Cengiz Öztireli and Olga Sorkine-Hornung},
  journal= {arXiv preprint arXiv:1906.04173},
  year   = {2019}
}

Comments

This version is contains camera-ready manuscript for SIGGRAPH Asia 2019

R2 v1 2026-06-23T09:49:15.573Z