{"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/graph-convolutional-label-noise-cleaner-train","title":"Graph Convolutional Label Noise Cleaner: Train a Plug-and-play Action Classifier for Anomaly Detection","arxiv_id":"1903.07256","date":"2019-03-18","proceeding":"CVPR 2019 6","authors":["Jia-Xing Zhong","Nannan Li","Weijie Kong","Shan Liu","Thomas H. Li","Ge Li"],"abstract":"Video anomaly detection under weak labels is formulated as a typical\nmultiple-instance learning problem in previous works. In this paper, we provide\na new perspective, i.e., a supervised learning task under noisy labels. In such\na viewpoint, as long as cleaning away label noise, we can directly apply fully\nsupervised action classifiers to weakly supervised anomaly detection, and take\nmaximum advantage of these well-developed classifiers. For this purpose, we\ndevise a graph convolutional network to correct noisy labels. Based upon\nfeature similarity and temporal consistency, our network propagates supervisory\nsignals from high-confidence snippets to low-confidence ones. In this manner,\nthe network is capable of providing cleaned supervision for action classifiers.\nDuring the test phase, we only need to obtain snippet-wise predictions from the\naction classifier without any extra post-processing. Extensive experiments on 3\ndatasets at different scales with 2 types of action classifiers demonstrate the\nefficacy of our method. Remarkably, we obtain the frame-level AUC score of\n82.12% on UCF-Crime.","url_abs":"http://arxiv.org/abs/1903.07256v1","url_pdf":"http://arxiv.org/pdf/1903.07256v1.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":"graph-convolutional-label-noise-cleaner-train","repo_url":"https://github.com/jx-zhong-for-academic-purpose/GCN-Anomaly-Detection","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"pytorch","reach":{"status":"ok"}}],"tasks":[{"task_slug":"anomaly-detection","task_name":"Anomaly Detection"},{"task_slug":"anomaly-detection-in-surveillance-videos","task_name":"Anomaly Detection In Surveillance Videos"},{"task_slug":"multiple-instance-learning","task_name":"Multiple Instance Learning"},{"task_slug":"supervised-anomaly-detection","task_name":"Supervised Anomaly Detection"},{"task_slug":"video-anomaly-detection","task_name":"Video Anomaly Detection"},{"task_slug":"weakly-supervised-anomaly-detection","task_name":"Weakly-supervised Anomaly Detection"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/anomaly-detection-in-surveillance-videos-on-1","task":"Anomaly Detection In Surveillance Videos","dataset":"ShanghaiTech Weakly Supervised","model":"GCN-Anomaly","rank_in_archive_order":12,"of":12,"metrics":{"AUC-ROC":"84.44"},"uses_additional_data":false},{"leaderboard":"/sota/anomaly-detection-in-surveillance-videos-on-3","task":"Anomaly Detection In Surveillance Videos","dataset":"UCSD Peds2","model":"GCN-Anomaly","rank_in_archive_order":6,"of":6,"metrics":{"AUC":"93.2"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1903.07256","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}