{"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/depth-and-dof-cues-make-a-better-defocus-blur","title":"Depth and DOF Cues Make A Better Defocus Blur Detector","arxiv_id":"2306.11334","date":"2023-06-20","proceeding":null,"authors":["Yuxin Jin","Ming Qian","Jincheng Xiong","Nan Xue","Gui-Song Xia"],"abstract":"Defocus blur detection (DBD) separates in-focus and out-of-focus regions in an image. Previous approaches mistakenly mistook homogeneous areas in focus for defocus blur regions, likely due to not considering the internal factors that cause defocus blur. Inspired by the law of depth, depth of field (DOF), and defocus, we propose an approach called D-DFFNet, which incorporates depth and DOF cues in an implicit manner. This allows the model to understand the defocus phenomenon in a more natural way. Our method proposes a depth feature distillation strategy to obtain depth knowledge from a pre-trained monocular depth estimation model and uses a DOF-edge loss to understand the relationship between DOF and depth. Our approach outperforms state-of-the-art methods on public benchmarks and a newly collected large benchmark dataset, EBD. Source codes and EBD dataset are available at: https:github.com/yuxinjin-whu/D-DFFNet.","url_abs":"https://arxiv.org/abs/2306.11334v1","url_pdf":"https://arxiv.org/pdf/2306.11334v1.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":"depth-and-dof-cues-make-a-better-defocus-blur","repo_url":"https://github.com/yuxinjin-whu/d-dffnet","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"defocus-blur-detection","task_name":"Defocus Blur Detection"},{"task_slug":"depth-estimation","task_name":"Depth Estimation"},{"task_slug":"monocular-depth-estimation","task_name":"Monocular Depth Estimation"}],"methods":[{"method_slug":"focus","method_name":"Focus"},{"method_slug":"knowledge-distillation","method_name":"Knowledge Distillation"}],"datasets_introduced":[{"slug":"ebd","name":"EBD","full_name":""}],"methods_introduced":[],"results":[{"leaderboard":"/sota/defocus-blur-detection-on-ctcug","task":"Defocus Blur Detection","dataset":"CTCUG","model":"D-DFFNet","rank_in_archive_order":1,"of":1,"metrics":{"IoU":"0.878","MAE":"0.074"},"uses_additional_data":false},{"leaderboard":"/sota/defocus-blur-detection-on-cuhk","task":"Defocus Blur Detection","dataset":"CUHK","model":"D-DFFNet","rank_in_archive_order":1,"of":2,"metrics":{"MAE":"0.036"},"uses_additional_data":false},{"leaderboard":"/sota/defocus-blur-detection-on-ebd","task":"Defocus Blur Detection","dataset":"EBD","model":"D-DFFNet","rank_in_archive_order":1,"of":1,"metrics":{"MAE":"0.084"},"uses_additional_data":false}],"syntology":{"syntology_url":"https://syntology.ai/paper/2306.11334","atlas_url":"https://app.syntology.ai/?focus=2306.11334","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}