{"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/linear-disentangled-representation-learning","title":"Linear Disentangled Representation Learning for Facial Actions","arxiv_id":"1701.03102","date":"2017-01-11","proceeding":null,"authors":["Xiang Xiang","Trac. D. Tran"],"abstract":"Limited annotated data available for the recognition of facial expression and\naction units embarrasses the training of deep networks, which can learn\ndisentangled invariant features. However, a linear model with just several\nparameters normally is not demanding in terms of training data. In this paper,\nwe propose an elegant linear model to untangle confounding factors in\nchallenging realistic multichannel signals such as 2D face videos. The simple\nyet powerful model does not rely on huge training data and is natural for\nrecognizing facial actions without explicitly disentangling the identity. Base\non well-understood intuitive linear models such as Sparse Representation based\nClassification (SRC), previous attempts require a prepossessing of explicit\ndecoupling which is practically inexact. Instead, we exploit the low-rank\nproperty across frames to subtract the underlying neutral faces which are\nmodeled jointly with sparse representation on the action components with group\nsparsity enforced. On the extended Cohn-Kanade dataset (CK+), our one-shot\nautomatic method on raw face videos performs as competitive as SRC applied on\nmanually prepared action components and performs even better than SRC in terms\nof true positive rate. We apply the model to the even more challenging task of\nfacial action unit recognition, verified on the MPI Face Video Database\n(MPI-VDB) achieving a decent performance. All the programs and data have been\nmade publicly available.","url_abs":"http://arxiv.org/abs/1701.03102v1","url_pdf":"http://arxiv.org/pdf/1701.03102v1.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":"linear-disentangled-representation-learning","repo_url":"https://github.com/eglxiang/FacialAU","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"none","reach":null},{"paper_slug":"linear-disentangled-representation-learning","repo_url":"https://github.com/eglxiang/icassp15_emotion","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"none","reach":null}],"tasks":[{"task_slug":"facial-action-unit-detection","task_name":"Facial Action Unit Detection"},{"task_slug":"representation-learning","task_name":"Representation Learning"},{"task_slug":"sparse-representation-based-classification","task_name":"Sparse Representation-based Classification"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}