Typicality-based point OOD detection with contrastive learning

Nawid Keshtmand, Raul Santos-Rodriguez, Jonathan Lawry
Proceedings of the 5th Northern Lights Deep Learning Conference ({NLDL}), PMLR 233:120-129, 2024.

Abstract

Typicality-based inference methods for OOD detection find a typical value (often the mean value) of a model statistic from the training data and then flag test points as anomalous if the model statistic of the test data point deviates significantly from the typical value. These methods are effective for detecting a group of OOD data points when OOD data points are labeled into groups, but ineffective for the detection of individual OOD data points. In this paper, we extend typicality-based inference to be effective for point OOD detection by utilizing latent features learned from contrastive learning and then obtaining the nearest neighbors of a test data point to provide additional context used for point OOD detection. The typicality-based inference approach is shown to improve point OOD detection relative to several benchmarks.

Cite this Paper

BibTeX
@InProceedings{pmlr-v233-keshtmand24a, title = {Typicality-based point {OOD} detection with contrastive learning}, author = {Keshtmand, Nawid and Santos-Rodriguez, Raul and Lawry, Jonathan}, booktitle = {Proceedings of the 5th Northern Lights Deep Learning Conference ({NLDL})}, pages = {120--129}, year = {2024}, editor = {Lutchyn, Tetiana and Ramírez Rivera, Adín and Ricaud, Benjamin}, volume = {233}, series = {Proceedings of Machine Learning Research}, month = {09--11 Jan}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v233/keshtmand24a/keshtmand24a.pdf}, url = {https://proceedings.mlr.press/v233/keshtmand24a.html}, abstract = {Typicality-based inference methods for OOD detection find a typical value (often the mean value) of a model statistic from the training data and then flag test points as anomalous if the model statistic of the test data point deviates significantly from the typical value. These methods are effective for detecting a group of OOD data points when OOD data points are labeled into groups, but ineffective for the detection of individual OOD data points. In this paper, we extend typicality-based inference to be effective for point OOD detection by utilizing latent features learned from contrastive learning and then obtaining the nearest neighbors of a test data point to provide additional context used for point OOD detection. The typicality-based inference approach is shown to improve point OOD detection relative to several benchmarks.} }
Endnote
%0 Conference Paper %T Typicality-based point OOD detection with contrastive learning %A Nawid Keshtmand %A Raul Santos-Rodriguez %A Jonathan Lawry %B Proceedings of the 5th Northern Lights Deep Learning Conference ({NLDL}) %C Proceedings of Machine Learning Research %D 2024 %E Tetiana Lutchyn %E Adín Ramírez Rivera %E Benjamin Ricaud %F pmlr-v233-keshtmand24a %I PMLR %P 120--129 %U https://proceedings.mlr.press/v233/keshtmand24a.html %V 233 %X Typicality-based inference methods for OOD detection find a typical value (often the mean value) of a model statistic from the training data and then flag test points as anomalous if the model statistic of the test data point deviates significantly from the typical value. These methods are effective for detecting a group of OOD data points when OOD data points are labeled into groups, but ineffective for the detection of individual OOD data points. In this paper, we extend typicality-based inference to be effective for point OOD detection by utilizing latent features learned from contrastive learning and then obtaining the nearest neighbors of a test data point to provide additional context used for point OOD detection. The typicality-based inference approach is shown to improve point OOD detection relative to several benchmarks.
APA
Keshtmand, N., Santos-Rodriguez, R. & Lawry, J.. (2024). Typicality-based point OOD detection with contrastive learning. Proceedings of the 5th Northern Lights Deep Learning Conference ({NLDL}), in Proceedings of Machine Learning Research 233:120-129 Available from https://proceedings.mlr.press/v233/keshtmand24a.html.

Related Material