PyDIff
The PyDiff project uses pyramid diffusion models for enhancing low-light images, delivering the best quantitative results compared to existing methods. Implemented within the PyDiff framework, it improves PSNR and SSIM scores. Instructions for installation, along with links to the LOL dataset and pretrained model, are provided. It offers flexible training options for multi-GPU and single GPU modes, suitable for custom low-level tasks. Built on the BasicSR framework, it offers advanced solutions in image enhancement.
