Reproduce “ENVIDR: Implicit Differentiable Renderer with Neural Environment Lighting” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Ruofan Liang, Huiting Chen, Chunlin Li, Fan Chen, Selvakumar Panneer, Nandita Vijaykumar
ENVIDR separates learned environment lighting, diffuse color, and specular reflection within an SDF-based scene model for reconstructing shiny objects from multiview images. It trains renderer components on physically based rendered examples, then uses them to represent scene appearance and enable relighting. On the Shiny Blender helmet, it reports normal MAE of 1.66° versus Ref-NeRF’s 29.48°, though results vary across benchmark metrics and scenes.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.