{"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-detection-with-end-to-end-integration-of","title":"Face Detection with End-to-End Integration of a ConvNet and a 3D Model","arxiv_id":"1606.00850","date":"2016-06-02","proceeding":null,"authors":["Yunzhu Li","Benyuan Sun","Tianfu Wu","Yizhou Wang"],"abstract":"This paper presents a method for face detection in the wild, which integrates\na ConvNet and a 3D mean face model in an end-to-end multi-task discriminative\nlearning framework. The 3D mean face model is predefined and fixed (e.g., we\nused the one provided in the AFLW dataset). The ConvNet consists of two\ncomponents: (i) The face pro- posal component computes face bounding box\nproposals via estimating facial key-points and the 3D transformation (rotation\nand translation) parameters for each predicted key-point w.r.t. the 3D mean\nface model. (ii) The face verification component computes detection results by\nprun- ing and refining proposals based on facial key-points based configuration\npooling. The proposed method addresses two issues in adapting state- of-the-art\ngeneric object detection ConvNets (e.g., faster R-CNN) for face detection: (i)\nOne is to eliminate the heuristic design of prede- fined anchor boxes in the\nregion proposals network (RPN) by exploit- ing a 3D mean face model. (ii) The\nother is to replace the generic RoI (Region-of-Interest) pooling layer with a\nconfiguration pooling layer to respect underlying object structures. The\nmulti-task loss consists of three terms: the classification Softmax loss and\nthe location smooth l1 -losses [14] of both the facial key-points and the face\nbounding boxes. In ex- periments, our ConvNet is trained on the AFLW dataset\nonly and tested on the FDDB benchmark with fine-tuning and on the AFW benchmark\nwithout fine-tuning. The proposed method obtains very competitive\nstate-of-the-art performance in the two benchmarks.","url_abs":"http://arxiv.org/abs/1606.00850v3","url_pdf":"http://arxiv.org/pdf/1606.00850v3.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-detection-with-end-to-end-integration-of","repo_url":"https://github.com/tfwu/FaceDetection-ConvNet-3D","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":null},{"paper_slug":"face-detection-with-end-to-end-integration-of","repo_url":"https://github.com/MindSpore-paper-code-3/code3/tree/main/FaceDetection","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"mindspore","reach":null},{"paper_slug":"face-detection-with-end-to-end-integration-of","repo_url":"https://github.com/MindSpore-paper-code-3/code7/tree/main/FaceDetection","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"mindspore","reach":null},{"paper_slug":"face-detection-with-end-to-end-integration-of","repo_url":"https://github.com/MindSpore-paper-code-3/code8/tree/main/FaceDetection","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"mindspore","reach":null},{"paper_slug":"face-detection-with-end-to-end-integration-of","repo_url":"https://github.com/Recognito-Vision/Android-FaceRecognition-FaceLivenessDetection","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"none","reach":{"status":"ok"}}],"tasks":[{"task_slug":"face-detection","task_name":"Face Detection"},{"task_slug":"face-model","task_name":"Face Model"},{"task_slug":"face-verification","task_name":"Face Verification"},{"task_slug":"object-detection","task_name":"Object Detection"},{"task_slug":"object-detection-1","task_name":"object-detection"}],"methods":[{"method_slug":"softmax","method_name":"Softmax"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/face-detection-on-annotated-faces-in-the-wild","task":"Face Detection","dataset":"Annotated Faces in the Wild","model":"Conv3D","rank_in_archive_order":7,"of":7,"metrics":{"AP":"0.9597"},"uses_additional_data":false}],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}