{"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/seqface-make-full-use-of-sequence-information","title":"SeqFace: Make full use of sequence information for face recognition","arxiv_id":"1803.06524","date":"2018-03-17","proceeding":null,"authors":["Wei Hu","Yangyu Huang","Fan Zhang","Ruirui Li","Wei Li","Guodong Yuan"],"abstract":"Deep convolutional neural networks (CNNs) have greatly improved the Face\nRecognition (FR) performance in recent years. Almost all CNNs in FR are trained\non the carefully labeled datasets containing plenty of identities. However,\nsuch high-quality datasets are very expensive to collect, which restricts many\nresearchers to achieve state-of-the-art performance. In this paper, we propose\na framework, called SeqFace, for learning discriminative face features. Besides\na traditional identity training dataset, the designed SeqFace can train CNNs by\nusing an additional dataset which includes a large number of face sequences\ncollected from videos. Moreover, the label smoothing regularization (LSR) and a\nnew proposed discriminative sequence agent (DSA) loss are employed to enhance\ndiscrimination power of deep face features via making full use of the sequence\ndata. Our method achieves excellent performance on Labeled Faces in the Wild\n(LFW), YouTube Faces (YTF), only with a single ResNet. The code and models are\npublicly available on-line (https://github.com/huangyangyu/SeqFace).","url_abs":"http://arxiv.org/abs/1803.06524v2","url_pdf":"http://arxiv.org/pdf/1803.06524v2.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":"seqface-make-full-use-of-sequence-information","repo_url":"https://github.com/huangyangyu/SeqFace","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"none","reach":null}],"tasks":[{"task_slug":"face-recognition","task_name":"Face Recognition"},{"task_slug":"face-verification","task_name":"Face Verification"}],"methods":[{"method_slug":"1x1-convolution","method_name":"1x1 Convolution"},{"method_slug":"average-pooling","method_name":"Average Pooling"},{"method_slug":"batch-normalization","method_name":"Batch Normalization"},{"method_slug":"bottleneck-residual-block","method_name":"Bottleneck Residual Block"},{"method_slug":"convolution","method_name":"Convolution"},{"method_slug":"global-average-pooling","method_name":"Global Average Pooling"},{"method_slug":"kaiming-initialization","method_name":"Kaiming Initialization"},{"method_slug":"label-smoothing","method_name":"Label Smoothing"},{"method_slug":"max-pooling","method_name":"Max Pooling"},{"method_slug":"relu","method_name":"ReLU"},{"method_slug":"residual-block","method_name":"Residual Block"},{"method_slug":"residual-connection","method_name":"Residual Connection"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/face-verification-on-youtube-faces-db","task":"Face Verification","dataset":"YouTube Faces DB","model":"SeqFace, 1 ResNet-64","rank_in_archive_order":1,"of":12,"metrics":{"Accuracy":"98.12%"},"uses_additional_data":false}],"syntology":{"syntology_url":null,"atlas_url":"https://app.syntology.ai/?focus=1803.06524","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}