{"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/shuffle-and-learn-unsupervised-learning-using","title":"Shuffle and Learn: Unsupervised Learning using Temporal Order Verification","arxiv_id":"1603.08561","date":"2016-03-28","proceeding":null,"authors":["Ishan Misra","C. Lawrence Zitnick","Martial Hebert"],"abstract":"In this paper, we present an approach for learning a visual representation\nfrom the raw spatiotemporal signals in videos. Our representation is learned\nwithout supervision from semantic labels. We formulate our method as an\nunsupervised sequential verification task, i.e., we determine whether a\nsequence of frames from a video is in the correct temporal order. With this\nsimple task and no semantic labels, we learn a powerful visual representation\nusing a Convolutional Neural Network (CNN). The representation contains\ncomplementary information to that learned from supervised image datasets like\nImageNet. Qualitative results show that our method captures information that is\ntemporally varying, such as human pose. When used as pre-training for action\nrecognition, our method gives significant gains over learning without external\ndata on benchmark datasets like UCF101 and HMDB51. To demonstrate its\nsensitivity to human pose, we show results for pose estimation on the FLIC and\nMPII datasets that are competitive, or better than approaches using\nsignificantly more supervision. Our method can be combined with supervised\nrepresentations to provide an additional boost in accuracy.","url_abs":"http://arxiv.org/abs/1603.08561v2","url_pdf":"http://arxiv.org/pdf/1603.08561v2.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":"action-recognition-in-videos","task_name":"Action Recognition"},{"task_slug":"pose-estimation","task_name":"Pose Estimation"},{"task_slug":"self-supervised-action-recognition","task_name":"Self-Supervised Action Recognition"},{"task_slug":"action-recognition","task_name":"Temporal Action Localization"},{"task_slug":"video-alignment","task_name":"Video Alignment"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/self-supervised-action-recognition-on-hmdb51","task":"Self-Supervised Action Recognition","dataset":"HMDB51","model":"Shuffle and Learn (AlexNet)","rank_in_archive_order":48,"of":48,"metrics":{"Frozen":"false","Pre-Training Dataset":"UCF101","Top-1 Accuracy":"19.8"},"uses_additional_data":false},{"leaderboard":"/sota/self-supervised-action-recognition-on-ucf101","task":"Self-Supervised Action Recognition","dataset":"UCF101","model":"Shuffle and Learn (AlexNet)","rank_in_archive_order":52,"of":53,"metrics":{"3-fold Accuracy":"50.9","Frozen":"false","Pre-Training Dataset":"UCF101"},"uses_additional_data":false}],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}