Reflectance Decomposition

Reflectance decomposition aims to separate an image's illumination and surface reflectance properties, a crucial step in various computer vision tasks. Current research focuses on improving the accuracy of this decomposition by addressing challenges like self-shadowing, inter-reflections, and non-Lambertian surfaces, often employing neural networks and physics-based models to disentangle these components. These advancements leverage techniques such as differentiable rendering, importance sampling, and spectral reflectance decomposition to achieve more robust and detailed material estimations. Improved reflectance decomposition has significant implications for applications like 3D face reconstruction, relighting, and material editing.