subeeshvasu/Awesome-Deblurring

Organizes deblurring research across multiple problem formulations—blind motion deblurring (both traditional and deep learning), non-blind deconvolution, multi-frame/video approaches, and depth-aware methods—with curated benchmark datasets and evaluation resources. Covers the evolution from classical optimization-based techniques (sparse priors, kernel estimation) to modern deep learning solutions, enabling researchers to trace methodological advances. Includes links to reproducible implementations, official project pages, and standardized datasets for motion and defocus blur across single-image and multi-frame scenarios.

2,839 stars. No commits in the last 6 months.