{"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/image-super-resolution-with-cross-scale-non","title":"Image Super-Resolution with Cross-Scale Non-Local Attention and Exhaustive Self-Exemplars Mining","arxiv_id":"2006.01424","date":"2020-06-02","proceeding":"CVPR 2020 6","authors":["Yiqun Mei","Yuchen Fan","Yuqian Zhou","Lichao Huang","Thomas S. Huang","Humphrey Shi"],"abstract":"Deep convolution-based single image super-resolution (SISR) networks embrace the benefits of learning from large-scale external image resources for local recovery, yet most existing works have ignored the long-range feature-wise similarities in natural images. Some recent works have successfully leveraged this intrinsic feature correlation by exploring non-local attention modules. However, none of the current deep models have studied another inherent property of images: cross-scale feature correlation. In this paper, we propose the first Cross-Scale Non-Local (CS-NL) attention module with integration into a recurrent neural network. By combining the new CS-NL prior with local and in-scale non-local priors in a powerful recurrent fusion cell, we can find more cross-scale feature correlations within a single low-resolution (LR) image. The performance of SISR is significantly improved by exhaustively integrating all possible priors. Extensive experiments demonstrate the effectiveness of the proposed CS-NL module by setting new state-of-the-arts on multiple SISR benchmarks.","url_abs":"https://arxiv.org/abs/2006.01424v1","url_pdf":"https://arxiv.org/pdf/2006.01424v1.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":"image-super-resolution-with-cross-scale-non","repo_url":"https://github.com/SHI-Labs/Cross-Scale-Non-Local-Attention","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"image-super-resolution-with-cross-scale-non","repo_url":"https://github.com/Lornatang/CSNLN-PyTorch","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"Apache-2.0"}},{"paper_slug":"image-super-resolution-with-cross-scale-non","repo_url":"https://github.com/nSamsow/CSNLN-new","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"feature-correlation","task_name":"Feature Correlation"},{"task_slug":"image-super-resolution","task_name":"Image Super-Resolution"},{"task_slug":"super-resolution","task_name":"Super-Resolution"}],"methods":[{"method_slug":"cross-scale-non-local-attention","method_name":"Cross-Scale Non-Local Attention"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/image-super-resolution-on-bsd100-2x-upscaling","task":"Image Super-Resolution","dataset":"BSD100 - 2x upscaling","model":"CSNLN","rank_in_archive_order":17,"of":30,"metrics":{"PSNR":"32.4","SSIM":"0.9024"},"uses_additional_data":false},{"leaderboard":"/sota/image-super-resolution-on-bsd100-3x-upscaling","task":"Image Super-Resolution","dataset":"BSD100 - 3x upscaling","model":"CSNLN","rank_in_archive_order":12,"of":21,"metrics":{"PSNR":"29.33","SSIM":"0.8105"},"uses_additional_data":false},{"leaderboard":"/sota/image-super-resolution-on-bsd100-4x-upscaling","task":"Image Super-Resolution","dataset":"BSD100 - 4x upscaling","model":"CSNLN","rank_in_archive_order":19,"of":71,"metrics":{"PSNR":"27.8","SSIM":"0.7439"},"uses_additional_data":false},{"leaderboard":"/sota/image-super-resolution-on-manga109-2x","task":"Image Super-Resolution","dataset":"Manga109 - 2x upscaling","model":"CSNLN","rank_in_archive_order":13,"of":21,"metrics":{"PSNR":"39.37","SSIM":"0.9785"},"uses_additional_data":false},{"leaderboard":"/sota/image-super-resolution-on-manga109-3x","task":"Image Super-Resolution","dataset":"Manga109 - 3x upscaling","model":"CSNLN","rank_in_archive_order":11,"of":17,"metrics":{"PSNR":"34.45","SSIM":"0.9502"},"uses_additional_data":false},{"leaderboard":"/sota/image-super-resolution-on-manga109-4x","task":"Image Super-Resolution","dataset":"Manga109 - 4x upscaling","model":"CSNLN","rank_in_archive_order":25,"of":50,"metrics":{"PSNR":"31.43","SSIM":"0.9201"},"uses_additional_data":false},{"leaderboard":"/sota/image-super-resolution-on-set14-2x-upscaling","task":"Image Super-Resolution","dataset":"Set14 - 2x upscaling","model":"CSNLN","rank_in_archive_order":16,"of":35,"metrics":{"PSNR":"34.12","SSIM":"0.9223"},"uses_additional_data":false},{"leaderboard":"/sota/image-super-resolution-on-set14-3x-upscaling","task":"Image Super-Resolution","dataset":"Set14 - 3x upscaling","model":"CSNLN","rank_in_archive_order":13,"of":24,"metrics":{"PSNR":"30.66","SSIM":"0.8482"},"uses_additional_data":false},{"leaderboard":"/sota/image-super-resolution-on-set14-4x-upscaling","task":"Image Super-Resolution","dataset":"Set14 - 4x upscaling","model":"CSNLN","rank_in_archive_order":34,"of":104,"metrics":{"PSNR":"28.95","SSIM":"0.7888"},"uses_additional_data":false},{"leaderboard":"/sota/image-super-resolution-on-set5-2x-upscaling","task":"Image Super-Resolution","dataset":"Set5 - 2x upscaling","model":"CSNLN","rank_in_archive_order":17,"of":41,"metrics":{"PSNR":"38.28","SSIM":"0.9616"},"uses_additional_data":false},{"leaderboard":"/sota/image-super-resolution-on-set5-3x-upscaling","task":"Image Super-Resolution","dataset":"Set5 - 3x upscaling","model":"CSNLN","rank_in_archive_order":15,"of":32,"metrics":{"PSNR":"34.74","SSIM":"0.9300"},"uses_additional_data":false},{"leaderboard":"/sota/image-super-resolution-on-urban100-2x","task":"Image Super-Resolution","dataset":"Urban100 - 2x upscaling","model":"CSNLN","rank_in_archive_order":14,"of":29,"metrics":{"PSNR":"33.25","SSIM":"0.9386"},"uses_additional_data":false},{"leaderboard":"/sota/image-super-resolution-on-urban100-3x","task":"Image Super-Resolution","dataset":"Urban100 - 3x upscaling","model":"CSNLN","rank_in_archive_order":13,"of":22,"metrics":{"PSNR":"29.13","SSIM":"0.8712"},"uses_additional_data":false},{"leaderboard":"/sota/image-super-resolution-on-urban100-4x","task":"Image Super-Resolution","dataset":"Urban100 - 4x upscaling","model":"CSNLN","rank_in_archive_order":19,"of":65,"metrics":{"PSNR":"27.22","SSIM":"0.8168"},"uses_additional_data":false}],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}