{"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/facenet2expnet-regularizing-a-deep-face","title":"FaceNet2ExpNet: Regularizing a Deep Face Recognition Net for Expression Recognition","arxiv_id":"1609.06591","date":"2016-09-21","proceeding":null,"authors":["Hui Ding","Shaohua Kevin Zhou","Rama Chellappa"],"abstract":"Relatively small data sets available for expression recognition research make\nthe training of deep networks for expression recognition very challenging.\nAlthough fine-tuning can partially alleviate the issue, the performance is\nstill below acceptable levels as the deep features probably contain redun- dant\ninformation from the pre-trained domain. In this paper, we present\nFaceNet2ExpNet, a novel idea to train an expression recognition network based\non static images. We first propose a new distribution function to model the\nhigh-level neurons of the expression network. Based on this, a two-stage\ntraining algorithm is carefully designed. In the pre-training stage, we train\nthe convolutional layers of the expression net, regularized by the face net; In\nthe refining stage, we append fully- connected layers to the pre-trained\nconvolutional layers and train the whole network jointly. Visualization shows\nthat the model trained with our method captures improved high-level expression\nsemantics. Evaluations on four public expression databases, CK+, Oulu-CASIA,\nTFD, and SFEW demonstrate that our method achieves better results than\nstate-of-the-art.","url_abs":"http://arxiv.org/abs/1609.06591v2","url_pdf":"http://arxiv.org/pdf/1609.06591v2.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":"face-recognition","task_name":"Face Recognition"},{"task_slug":"facial-expression-recognition","task_name":"Facial Expression Recognition (FER)"},{"task_slug":"small-data","task_name":"Small Data Image Classification"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/facial-expression-recognition-on-ck","task":"Facial Expression Recognition (FER)","dataset":"CK+","model":"FN2EN","rank_in_archive_order":2,"of":7,"metrics":{"Accuracy (6 emotion)":"98.6","Accuracy (7 emotion)":"-","Accuracy (8 emotion)":"96.8"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1609.06591","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}