The paper "Recasting Regional Lighting for Shadow Removal" addresses the challenge of removing shadows from images by focusing on the degradation of object textures in shadow regions. Traditional methods often fail to fully recover attenuated textures, even when enhancing local illumination. The authors propose a novel approach that explicitly estimates the illumination and reflectance layers of shadow regions using a shadow-aware decomposition network. This network is followed by a bilateral correction network that recasts the lighting and restores the textures conditionally. The bilateral correction network includes two modules: a local lighting correction (LLC) module and an illumination-guided texture restoration (IGTR) module. The LLC module uses a diffusion model to iteratively correct the local lighting, while the IGTR module enhances texture consistency through scale-adaptive feature consistency enhancement. The method is evaluated on three benchmarks, demonstrating superior performance compared to existing state-of-the-art methods. Key contributions include the introduction of the shadow-aware decomposition network and the bilateral correction network, as well as the manual annotation of shadow masks for fair evaluation.The paper "Recasting Regional Lighting for Shadow Removal" addresses the challenge of removing shadows from images by focusing on the degradation of object textures in shadow regions. Traditional methods often fail to fully recover attenuated textures, even when enhancing local illumination. The authors propose a novel approach that explicitly estimates the illumination and reflectance layers of shadow regions using a shadow-aware decomposition network. This network is followed by a bilateral correction network that recasts the lighting and restores the textures conditionally. The bilateral correction network includes two modules: a local lighting correction (LLC) module and an illumination-guided texture restoration (IGTR) module. The LLC module uses a diffusion model to iteratively correct the local lighting, while the IGTR module enhances texture consistency through scale-adaptive feature consistency enhancement. The method is evaluated on three benchmarks, demonstrating superior performance compared to existing state-of-the-art methods. Key contributions include the introduction of the shadow-aware decomposition network and the bilateral correction network, as well as the manual annotation of shadow masks for fair evaluation.