AI & ML New Capability

First unified pipeline to reconstruct complete geometry, materials, and lighting from sparse views in under one second.

March 23, 2026

Original Paper

ReLi3D: Relightable Multi-view 3D Reconstruction with Disentangled Illumination

Jan-Niklas Dihlmann, Mark Boss, Simon Donne, Andreas Engelhardt, Hendrik P.A. Lensch, Varun Jampani

arXiv · 2603.19753

The Takeaway

Current relightable 3D reconstruction is a fragmented process requiring minutes or hours of optimization. ReLi3D uses a transformer-based feed-forward approach to generate high-quality, relightable 3D assets nearly instantaneously, enabling real-time 3D asset generation from just a few images.

From the abstract

Reconstructing 3D assets from images has long required separate pipelines for geometry reconstruction, material estimation, and illumination recovery, each with distinct limitations and computational overhead. We present ReLi3D, the first unified end-to-end pipeline that simultaneously reconstructs complete 3D geometry, spatially-varying physically-based materials, and environment illumination from sparse multi-view images in under one second. Our key insight is that multi-view constraints can d