{"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/face-magnet-magnifying-feature-maps-to-detect","title":"Face-MagNet: Magnifying Feature Maps to Detect Small Faces","arxiv_id":"1803.05258","date":"2018-03-14","proceeding":null,"authors":["Pouya Samangouei","Mahyar Najibi","Larry Davis","Rama Chellappa"],"abstract":"In this paper, we introduce the Face Magnifier Network (Face-MageNet), a face\ndetector based on the Faster-RCNN framework which enables the flow of\ndiscriminative information of small scale faces to the classifier without any\nskip or residual connections. To achieve this, Face-MagNet deploys a set of\nConvTranspose, also known as deconvolution, layers in the Region Proposal\nNetwork (RPN) and another set before the Region of Interest (RoI) pooling layer\nto facilitate detection of finer faces. In addition, we also design, train, and\nevaluate three other well-tuned architectures that represent the conventional\nsolutions to the scale problem: context pooling, skip connections, and scale\npartitioning. Each of these three networks achieves comparable results to the\nstate-of-the-art face detectors. With extensive experiments, we show that\nFace-MagNet based on a VGG16 architecture achieves better results than the\nrecently proposed ResNet101-based HR method on the task of face detection on\nWIDER dataset and also achieves similar results on the hard set as our other\nmethod SSH.","url_abs":"http://arxiv.org/abs/1803.05258v1","url_pdf":"http://arxiv.org/pdf/1803.05258v1.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":"face-magnet-magnifying-feature-maps-to-detect","repo_url":"https://github.com/po0ya/face-magnet","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"caffe2","reach":null}],"tasks":[{"task_slug":"face-detection","task_name":"Face Detection"},{"task_slug":"region-proposal","task_name":"Region Proposal"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}