{"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/semi-orthogonal-embedding-for-efficient","title":"Semi-orthogonal Embedding for Efficient Unsupervised Anomaly Segmentation","arxiv_id":"2105.14737","date":"2021-05-31","proceeding":null,"authors":["Jin-Hwa Kim","Do-Hyeong Kim","Saehoon Yi","Taehoon Lee"],"abstract":"We present the efficiency of semi-orthogonal embedding for unsupervised anomaly segmentation. The multi-scale features from pre-trained CNNs are recently used for the localized Mahalanobis distances with significant performance. However, the increased feature size is problematic to scale up to the bigger CNNs, since it requires the batch-inverse of multi-dimensional covariance tensor. Here, we generalize an ad-hoc method, random feature selection, into semi-orthogonal embedding for robust approximation, cubically reducing the computational cost for the inverse of multi-dimensional covariance tensor. With the scrutiny of ablation studies, the proposed method achieves a new state-of-the-art with significant margins for the MVTec AD, KolektorSDD, KolektorSDD2, and mSTC datasets. The theoretical and empirical analyses offer insights and verification of our straightforward yet cost-effective approach.","url_abs":"https://arxiv.org/abs/2105.14737v1","url_pdf":"https://arxiv.org/pdf/2105.14737v1.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":"semi-orthogonal-embedding-for-efficient","repo_url":"https://github.com/jnhwkim/orthoad","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"pytorch","reach":null},{"paper_slug":"semi-orthogonal-embedding-for-efficient","repo_url":"https://github.com/Pangoraw/SemiOrthogonal","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"semi-orthogonal-embedding-for-efficient","repo_url":"https://github.com/Ultranity/Anomaly.Paddle","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"paddle","reach":{"status":"ok","spdx":"Apache-2.0"}}],"tasks":[{"task_slug":"anomaly-detection","task_name":"Anomaly Detection"},{"task_slug":"anomaly-segmentation","task_name":"Anomaly Segmentation"},{"task_slug":"unsupervised-anomaly-detection","task_name":"Unsupervised Anomaly Detection"},{"task_slug":"feature-selection","task_name":"feature selection"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/anomaly-detection-on-mvtec-ad","task":"Anomaly Detection","dataset":"MVTec AD","model":"Semi-orthogonal","rank_in_archive_order":135,"of":148,"metrics":{"Segmentation AUROC":"98.2"},"uses_additional_data":false},{"leaderboard":"/sota/unsupervised-anomaly-detection-on-kolektorsdd","task":"Unsupervised Anomaly Detection","dataset":"KolektorSDD","model":"Semi-orthogonal","rank_in_archive_order":1,"of":1,"metrics":{"Segmentation AUROC":"96.0"},"uses_additional_data":true},{"leaderboard":"/sota/unsupervised-anomaly-detection-on","task":"Unsupervised Anomaly Detection","dataset":"KolektorSDD2","model":"Semi-orthogonal","rank_in_archive_order":3,"of":3,"metrics":{"Segmentation AUROC":"98.1"},"uses_additional_data":true}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2105.14737","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}