{"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/sfa-small-faces-attention-face-detector","title":"SFA: Small Faces Attention Face Detector","arxiv_id":"1812.08402","date":"2018-12-20","proceeding":null,"authors":["Shi Luo","Xiongfei Li","Rui Zhu","Xiaoli Zhang"],"abstract":"In recent year, tremendous strides have been made in face detection thanks to\ndeep learning. However, most published face detectors deteriorate dramatically\nas the faces become smaller. In this paper, we present the Small Faces\nAttention (SFA) face detector to better detect faces with small scale. First,\nwe propose a new scale-invariant face detection architecture which pays more\nattention to small faces, including 4-branch detection architecture and small\nfaces sensitive anchor design. Second, feature maps fusion strategy is applied\nin SFA by partially combining high-level features into low-level features to\nfurther improve the ability of finding hard faces. Third, we use multi-scale\ntraining and testing strategy to enhance face detection performance in\npractice. Comprehensive experiments show that SFA significantly improves face\ndetection performance, especially on small faces. Our real-time SFA face\ndetector can run at 5 FPS on a single GPU as well as maintain high performance.\nBesides, our final SFA face detector achieves state-of-the-art detection\nperformance on challenging face detection benchmarks, including WIDER FACE and\nFDDB datasets, with competitive runtime speed. Both our code and models will be\navailable to the research community.","url_abs":"http://arxiv.org/abs/1812.08402v1","url_pdf":"http://arxiv.org/pdf/1812.08402v1.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":"sfa-small-faces-attention-face-detector","repo_url":"https://github.com/shiluo1990/SFA","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"none","reach":null},{"paper_slug":"sfa-small-faces-attention-face-detector","repo_url":"https://github.com/yangyucheng000/papercode-2/tree/main/SFAT-MindSpore-main","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"mindspore","reach":null}],"tasks":[{"task_slug":"face-detection","task_name":"Face Detection"},{"task_slug":null,"task_name":"GPU"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1812.08402","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}