{"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/motion-aware-double-attention-network-for","title":"Motion Aware Double Attention Network for Dynamic Scene Deblurring","arxiv_id":null,"date":"2020-06-19","proceeding":"Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops 2020 6","authors":["Dan Yang","Mehmet Yamac"],"abstract":"Motion deblurring in dynamic scenes is a challenging\r\ntask when the blurring is caused by one or a combination of\r\nvarious reasons such as moving objects, camera movement,\r\netc. Since event cameras can detect changes in intensity\r\nwith a low latency, necessary motion information is inherently captured in event data, which could be quite useful\r\nfor deblurring standard camera images. The degradation\r\nintensity does not show homogeneity across an image due\r\nto factors like object depth, speed, etc. We propose a twobranch network structure, Motion Aware Double Attention\r\nNetwork (MADANet), that pays special attention to areas\r\nwith high blur. As part of the network, event data is first\r\nused by the high blur region segmentation module that creates a probability-like score for areas exhibiting high relative motion to the camera. Then, the event data is also\r\ninjected to feature maps in the main body, where there is\r\na second attention mechanism available for each branch.\r\nThe effective usage of event data and two-level attention\r\nmechanisms makes the network very compact. During the\r\nexperiment, it was shown that the proposed network could\r\nachieve state-of-the-art performance not only on the benchmark dataset from GoPro, but also on two newly collected\r\ndatasets, one of which contains real event data","url_abs":"https://openaccess.thecvf.com/content/CVPR2022W/NTIRE/html/Yang_Motion_Aware_Double_Attention_Network_for_Dynamic_Scene_Deblurring_CVPRW_2022_paper.html","url_pdf":"https://openaccess.thecvf.com/content/CVPR2022W/NTIRE/papers/Yang_Motion_Aware_Double_Attention_Network_for_Dynamic_Scene_Deblurring_CVPRW_2022_paper.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":[],"tasks":[{"task_slug":"deblurring","task_name":"Deblurring"},{"task_slug":"image-deblurring","task_name":"Image Deblurring"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/deblurring-on-gopro","task":"Deblurring","dataset":"GoPro","model":"MADANet","rank_in_archive_order":11,"of":56,"metrics":{"PSNR":"33.84","SSIM":"0.964"},"uses_additional_data":true},{"leaderboard":"/sota/image-deblurring-on-gopro","task":"Image Deblurring","dataset":"GoPro","model":"MADANet","rank_in_archive_order":12,"of":55,"metrics":{"PSNR":"33.84","Params (M)":"16.9","SSIM":"0.964"},"uses_additional_data":true}],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}