{"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/gated-context-aggregation-network-for-image","title":"Gated Context Aggregation Network for Image Dehazing and Deraining","arxiv_id":"1811.08747","date":"2018-11-21","proceeding":null,"authors":["Dongdong Chen","Mingming He","Qingnan Fan","Jing Liao","Liheng Zhang","Dongdong Hou","Lu Yuan","Gang Hua"],"abstract":"Image dehazing aims to recover the uncorrupted content from a hazy image.\nInstead of leveraging traditional low-level or handcrafted image priors as the\nrestoration constraints, e.g., dark channels and increased contrast, we propose\nan end-to-end gated context aggregation network to directly restore the final\nhaze-free image. In this network, we adopt the latest smoothed dilation\ntechnique to help remove the gridding artifacts caused by the widely-used\ndilated convolution with negligible extra parameters, and leverage a gated\nsub-network to fuse the features from different levels. Extensive experiments\ndemonstrate that our method can surpass previous state-of-the-art methods by a\nlarge margin both quantitatively and qualitatively. In addition, to demonstrate\nthe generality of the proposed method, we further apply it to the image\nderaining task, which also achieves the state-of-the-art performance. Code has\nbeen made available at https://github.com/cddlyf/GCANet.","url_abs":"http://arxiv.org/abs/1811.08747v2","url_pdf":"http://arxiv.org/pdf/1811.08747v2.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":"gated-context-aggregation-network-for-image","repo_url":"https://github.com/cddlyf/GCANet","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"pytorch","reach":{"status":"ok"}}],"tasks":[{"task_slug":"image-dehazing","task_name":"Image Dehazing"},{"task_slug":"rain-removal","task_name":"Rain Removal"}],"methods":[{"method_slug":"convolution","method_name":"Convolution"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/image-dehazing-on-rs-haze","task":"Image Dehazing","dataset":"RS-Haze","model":"GCANet","rank_in_archive_order":6,"of":7,"metrics":{"PSNR":"34.41","SSIM":"0.949"},"uses_additional_data":false},{"leaderboard":"/sota/image-dehazing-on-sots-indoor","task":"Image Dehazing","dataset":"SOTS Indoor","model":"GCANet","rank_in_archive_order":27,"of":34,"metrics":{"PSNR":"30.23","SSIM":"0.98"},"uses_additional_data":false},{"leaderboard":"/sota/rain-removal-on-did-mdn","task":"Rain Removal","dataset":"DID-MDN","model":"GCANet","rank_in_archive_order":2,"of":2,"metrics":{"PSNR":"31.68"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1811.08747","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}