{"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/group-based-sparse-representation-for-image","title":"Group-based Sparse Representation for Image Restoration","arxiv_id":"1405.3351","date":"2014-05-14","proceeding":null,"authors":["Jian Zhang","Debin Zhao","Wen Gao"],"abstract":"Traditional patch-based sparse representation modeling of natural images\nusually suffer from two problems. First, it has to solve a large-scale\noptimization problem with high computational complexity in dictionary learning.\nSecond, each patch is considered independently in dictionary learning and\nsparse coding, which ignores the relationship among patches, resulting in\ninaccurate sparse coding coefficients. In this paper, instead of using patch as\nthe basic unit of sparse representation, we exploit the concept of group as the\nbasic unit of sparse representation, which is composed of nonlocal patches with\nsimilar structures, and establish a novel sparse representation modeling of\nnatural images, called group-based sparse representation (GSR). The proposed\nGSR is able to sparsely represent natural images in the domain of group, which\nenforces the intrinsic local sparsity and nonlocal self-similarity of images\nsimultaneously in a unified framework. Moreover, an effective self-adaptive\ndictionary learning method for each group with low complexity is designed,\nrather than dictionary learning from natural images. To make GSR tractable and\nrobust, a split Bregman based technique is developed to solve the proposed\nGSR-driven minimization problem for image restoration efficiently. Extensive\nexperiments on image inpainting, image deblurring and image compressive sensing\nrecovery manifest that the proposed GSR modeling outperforms many current\nstate-of-the-art schemes in both PSNR and visual perception.","url_abs":"http://arxiv.org/abs/1405.3351v1","url_pdf":"http://arxiv.org/pdf/1405.3351v1.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":"group-based-sparse-representation-for-image","repo_url":"https://github.com/jianzhangcs/GSR","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":null}],"tasks":[{"task_slug":"compressive-sensing","task_name":"Compressive Sensing"},{"task_slug":"deblurring","task_name":"Deblurring"},{"task_slug":"dictionary-learning","task_name":"Dictionary Learning"},{"task_slug":"image-deblurring","task_name":"Image Deblurring"},{"task_slug":"image-inpainting","task_name":"Image Inpainting"},{"task_slug":"image-restoration","task_name":"Image Restoration"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1405.3351","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}