CVPR2025

Lux Post Facto: Learning Portrait Performance Relighting with Conditional Video Diffusion and a Hybrid Dataset

Yiqun Mei, Mingming He, Li Ma, Julien Philip, Wenqi Xian, David M. George, Xueming Yu, Gabriel Dedic, Ahmet Levent Tasel, Ning Yu, Vishal M. Patel, Paul E. Debevec

Abstract

Figure 1 . Lux Post Facto offers portrait relighting as a simple post-production process. Users can edit the lighting of portrait images (first row) and videos (second row) with high fidelity using any HDR map. Our method is temporally stable and highly photorealistic.