Image Deraining

Image deraining aims to computationally remove rain streaks and raindrops from images, improving visibility and enabling accurate analysis in various applications like autonomous driving and aerial inspection. Current research focuses on developing sophisticated deep learning models, including transformers and convolutional neural networks, often incorporating multi-scale processing, frequency domain analysis, and contrastive learning techniques to effectively separate rain artifacts from the underlying scene. These advancements are crucial for enhancing the reliability of computer vision systems operating in adverse weather conditions and improving the quality of images for various downstream tasks.