{"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/metric-learning-and-adaptive-boundary-for-out","title":"Metric Learning and Adaptive Boundary for Out-of-Domain Detection","arxiv_id":"2204.10849","date":"2022-04-22","proceeding":null,"authors":["Petr Lorenc","Tommaso Gargiani","Jan Pichl","Jakub Konrád","Petr Marek","Ondřej Kobza","Jan Šedivý"],"abstract":"Conversational agents are usually designed for closed-world environments. Unfortunately, users can behave unexpectedly. Based on the open-world environment, we often encounter the situation that the training and test data are sampled from different distributions. Then, data from different distributions are called out-of-domain (OOD). A robust conversational agent needs to react to these OOD utterances adequately. Thus, the importance of robust OOD detection is emphasized. Unfortunately, collecting OOD data is a challenging task. We have designed an OOD detection algorithm independent of OOD data that outperforms a wide range of current state-of-the-art algorithms on publicly available datasets. Our algorithm is based on a simple but efficient approach of combining metric learning with adaptive decision boundary. Furthermore, compared to other algorithms, we have found that our proposed algorithm has significantly improved OOD performance in a scenario with a lower number of classes while preserving the accuracy for in-domain (IND) classes.","url_abs":"https://arxiv.org/abs/2204.10849v1","url_pdf":"https://arxiv.org/pdf/2204.10849v1.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":"metric-learning-and-adaptive-boundary-for-out","repo_url":"https://github.com/tgargiani/adaptive-boundary","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":null}],"tasks":[{"task_slug":"metric-learning","task_name":"Metric Learning"},{"task_slug":"open-intent-detection","task_name":"Open Intent Detection"},{"task_slug":"ood-detection","task_name":"Out of Distribution (OOD) Detection"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/open-intent-detection-on-banking-77-50-known","task":"Open Intent Detection","dataset":"BANKING-77 (50% known)","model":"Metric learning + Adaptive Decision Boundary","rank_in_archive_order":1,"of":2,"metrics":{"1:1 Accuracy":"83.78","F1-score":"84.93"},"uses_additional_data":false},{"leaderboard":"/sota/open-intent-detection-on-banking-77-75-known","task":"Open Intent Detection","dataset":"BANKING-77 (75% known)","model":"Metric learning + Adaptive Decision Boundary","rank_in_archive_order":1,"of":2,"metrics":{"1:1 Accuracy":"84.4","F1-score":"88.39"},"uses_additional_data":false},{"leaderboard":"/sota/open-intent-detection-on-banking77-25-known","task":"Open Intent Detection","dataset":"BANKING77 (25%known)","model":"Metric learning + Adaptive Decision Boundary","rank_in_archive_order":1,"of":2,"metrics":{"1:1 Accuracy":"85.71","F1-score":"78.86"},"uses_additional_data":false},{"leaderboard":"/sota/open-intent-detection-on-oos-25-known","task":"Open Intent Detection","dataset":"OOS(25%known)","model":"Metric learning + Adaptive Decision Boundary","rank_in_archive_order":1,"of":2,"metrics":{"1:1 Accuracy":"91.81","F1-score":"85.9"},"uses_additional_data":false},{"leaderboard":"/sota/open-intent-detection-on-oos-50-known","task":"Open Intent Detection","dataset":"OOS(50%known)","model":"Metric learning + Adaptive Decision Boundary","rank_in_archive_order":1,"of":2,"metrics":{"1:1 Accuracy":"88.81","F1-score":"89.19"},"uses_additional_data":false},{"leaderboard":"/sota/open-intent-detection-on-oos-75-known","task":"Open Intent Detection","dataset":"OOS(75%known)","model":"Metric learning + Adaptive Decision Boundary","rank_in_archive_order":1,"of":2,"metrics":{"1:1 Accuracy":"88.54","F1-score":"92.21"},"uses_additional_data":false}],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}