EMLight: Lighting Estimation via Spherical Distribution Approximation

Illumination estimation from a single image is critical in 3D rendering and it has been investigated extensively the computer vision graphic research community. On other hand, existing works estimate illumination by either regressing light parameters or generating maps that are often hard to optimize tend produce inaccurate predictions. We propose Earth Mover’s Light (EMLight), an framework leverages regression network neural projector for accurate estimation. decompose map into spherical distribution, intensity ambient term, define as parameter task three components. Motivated Mover's distance, we design novel mover's loss guides regress distribution accurately taking advantage of subtleties distribution. Under guidance predicted synthesizes panoramic with realistic frequency. Extensive experiments show EMLight achieves generated relighting object embedding exhibits superior plausibility fidelity compared state-of-the-art methods.