VEnhancer is an official implementation of a research paper on generative video enhancement. It uses diffusion models to improve AI-generated videos through spatial super-resolution (increasing resolution), temporal super-resolution (frame interpolation), and general video refinement. The model is designed specifically to enhance videos produced by other AI video generation systems.