{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/the-little-engine-that-could-regularization","title":"The Little Engine that Could: Regularization by Denoising (RED)","arxiv_id":"1611.02862","date":"2016-11-09","proceeding":null,"authors":["Yaniv Romano","Michael Elad","Peyman Milanfar"],"abstract":"Removal of noise from an image is an extensively studied problem in image\nprocessing. Indeed, the recent advent of sophisticated and highly effective\ndenoising algorithms lead some to believe that existing methods are touching\nthe ceiling in terms of noise removal performance. Can we leverage this\nimpressive achievement to treat other tasks in image processing? Recent work\nhas answered this question positively, in the form of the Plug-and-Play Prior\n($P^3$) method, showing that any inverse problem can be handled by sequentially\napplying image denoising steps. This relies heavily on the ADMM optimization\ntechnique in order to obtain this chained denoising interpretation.\n  Is this the only way in which tasks in image processing can exploit the image\ndenoising engine? In this paper we provide an alternative, more powerful and\nmore flexible framework for achieving the same goal. As opposed to the $P^3$\nmethod, we offer Regularization by Denoising (RED): using the denoising engine\nin defining the regularization of the inverse problem. We propose an explicit\nimage-adaptive Laplacian-based regularization functional, making the overall\nobjective functional clearer and better defined. With a complete flexibility to\nchoose the iterative optimization procedure for minimizing the above\nfunctional, RED is capable of incorporating any image denoising algorithm,\ntreat general inverse problems very effectively, and is guaranteed to converge\nto the globally optimal result. We test this approach and demonstrate\nstate-of-the-art results in the image deblurring and super-resolution problems.","url_abs":"http://arxiv.org/abs/1611.02862v3","url_pdf":"http://arxiv.org/pdf/1611.02862v3.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"the-little-engine-that-could-regularization","repo_url":"https://github.com/google/RED","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok","spdx":"Apache-2.0"}},{"paper_slug":"the-little-engine-that-could-regularization","repo_url":"https://github.com/happyhongt/Acceleration-of-RED-via-Vector-Extrapolation","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok"}}],"tasks":[{"task_slug":"deblurring","task_name":"Deblurring"},{"task_slug":"denoising","task_name":"Denoising"},{"task_slug":"image-deblurring","task_name":"Image Deblurring"},{"task_slug":"image-denoising","task_name":"Image Denoising"},{"task_slug":"super-resolution","task_name":"Super-Resolution"}],"methods":[{"method_slug":"admm","method_name":"ADMM"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1611.02862","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}