Image Denoiser

Image denoising aims to remove noise from images while preserving important details, a crucial preprocessing step for many computer vision tasks. Current research focuses on developing deep learning-based denoisers, employing architectures like convolutional neural networks (CNNs), transformers, and autoencoders, often incorporating techniques such as graph-based regularization or adversarial training to improve performance and robustness. These advancements are significantly impacting various fields, from medical imaging (improving diagnostic accuracy) to video processing (enhancing moment retrieval) and even improving the robustness of other machine learning models by acting as a pre-processing step.