Papers › Metric Learning and Adaptive Boundary for Out-of-Domain Detection

Metric Learning and Adaptive Boundary for Out-of-Domain Detection

22 Apr 2022arXiv:2204.10849archive 2025-07-28

Petr Lorenc, Tommaso Gargiani, Jan Pichl, Jakub Konrád, Petr Marek, Ondřej Kobza, Jan Šedivý

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.

PaperPDFCode

Code

tgargiani/adaptive-boundary officialmentioned in papermentioned on GitHubtf report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Metric LearningOpen Intent DetectionOut of Distribution (OOD) Detection

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Open Intent Detection BANKING-77 (50% known) Metric learning + Adaptive Decision Boundary 1:1 Accuracy 83.78 #1 of 2 Archive leaderboard report
Open Intent Detection BANKING-77 (50% known) Metric learning + Adaptive Decision Boundary F1-score 84.93 #1 of 2 Archive leaderboard report
Open Intent Detection BANKING-77 (75% known) Metric learning + Adaptive Decision Boundary 1:1 Accuracy 84.4 #1 of 2 Archive leaderboard report
Open Intent Detection BANKING-77 (75% known) Metric learning + Adaptive Decision Boundary F1-score 88.39 #1 of 2 Archive leaderboard report
Open Intent Detection BANKING77 (25%known) Metric learning + Adaptive Decision Boundary 1:1 Accuracy 85.71 #1 of 2 Archive leaderboard report
Open Intent Detection BANKING77 (25%known) Metric learning + Adaptive Decision Boundary F1-score 78.86 #1 of 2 Archive leaderboard report
Open Intent Detection OOS(25%known) Metric learning + Adaptive Decision Boundary 1:1 Accuracy 91.81 #1 of 2 Archive leaderboard report
Open Intent Detection OOS(25%known) Metric learning + Adaptive Decision Boundary F1-score 85.9 #1 of 2 Archive leaderboard report
Open Intent Detection OOS(50%known) Metric learning + Adaptive Decision Boundary 1:1 Accuracy 88.81 #1 of 2 Archive leaderboard report
Open Intent Detection OOS(50%known) Metric learning + Adaptive Decision Boundary F1-score 89.19 #1 of 2 Archive leaderboard report
Open Intent Detection OOS(75%known) Metric learning + Adaptive Decision Boundary 1:1 Accuracy 88.54 #1 of 2 Archive leaderboard report
Open Intent Detection OOS(75%known) Metric learning + Adaptive Decision Boundary F1-score 92.21 #1 of 2 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections