{"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/fast-and-interpretable-nonlocal-neural","title":"Fast and Interpretable Nonlocal Neural Networks for Image Denoising via Group-Sparse Convolutional Dictionary Learning","arxiv_id":"2306.01950","date":"2023-06-02","proceeding":null,"authors":["Nikola Janjušević","Amirhossein Khalilian-Gourtani","Adeen Flinker","Yao Wang"],"abstract":"Nonlocal self-similarity within natural images has become an increasingly popular prior in deep-learning models. Despite their successful image restoration performance, such models remain largely uninterpretable due to their black-box construction. Our previous studies have shown that interpretable construction of a fully convolutional denoiser (CDLNet), with performance on par with state-of-the-art black-box counterparts, is achievable by unrolling a dictionary learning algorithm. In this manuscript, we seek an interpretable construction of a convolutional network with a nonlocal self-similarity prior that performs on par with black-box nonlocal models. We show that such an architecture can be effectively achieved by upgrading the $\\ell 1$ sparsity prior of CDLNet to a weighted group-sparsity prior. From this formulation, we propose a novel sliding-window nonlocal operation, enabled by sparse array arithmetic. In addition to competitive performance with black-box nonlocal DNNs, we demonstrate the proposed sliding-window sparse attention enables inference speeds greater than an order of magnitude faster than its competitors.","url_abs":"https://arxiv.org/abs/2306.01950v1","url_pdf":"https://arxiv.org/pdf/2306.01950v1.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":"fast-and-interpretable-nonlocal-neural","repo_url":"https://github.com/nikopj/groupcdl-tip","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"none","reach":null}],"tasks":[{"task_slug":"denoising","task_name":"Denoising"},{"task_slug":"dictionary-learning","task_name":"Dictionary Learning"},{"task_slug":"grayscale-image-denoising","task_name":"Grayscale Image Denoising"},{"task_slug":"image-denoising","task_name":"Image Denoising"},{"task_slug":"image-restoration","task_name":"Image Restoration"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/grayscale-image-denoising-on-bsd68-sigma15","task":"Grayscale Image Denoising","dataset":"BSD68 sigma15","model":"GroupCDL","rank_in_archive_order":8,"of":16,"metrics":{"PSNR":"31.82"},"uses_additional_data":false},{"leaderboard":"/sota/grayscale-image-denoising-on-bsd68-sigma25","task":"Grayscale Image Denoising","dataset":"BSD68 sigma25","model":"GroupCDL","rank_in_archive_order":5,"of":16,"metrics":{"PSNR":"29.38"},"uses_additional_data":false},{"leaderboard":"/sota/grayscale-image-denoising-on-bsd68-sigma50","task":"Grayscale Image Denoising","dataset":"BSD68 sigma50","model":"GroupCDL","rank_in_archive_order":6,"of":15,"metrics":{"PSNR":"26.47"},"uses_additional_data":false}],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}