Interpretable Out-of-Distribution Detection using Pattern Identification

Romain Xu-Darme, Julien Girard-Satabin, Darryl Hond, Gabriele Incorvaia, Zakaria Chihani

Keywords: out-of-distribution detection, pattern detection, interpretable artificial intelligence, confidence, metric

TL;DR: We apply pattern detection to Out-of-Distribution detection on an extensive benchmark.

Abstract: Out-of-distribution (OoD) detection for data-based programs is a goal of paramount importance. Common approaches in the literature tend to train binary classifiers requiring inside-of-distribution (IoD) and OoD validation samples, and/or implement confidence metrics that are often abstract and therefore difficult to interpret. In this work, we propose to use the PARTICUL pattern identification algorithm in order to build more interpretable and robust OoD detectors for visual classifiers. Crucially, this approach does not require retraining the classifier and is tuned directly to the IoD dataset, making it applicable to domains where OoD does not have a clear definition. Moreover, pattern identification allows us to provide images from the IoD dataset as reference points to better explain our confidence scores. We illustrate the generalization abilities of our approach through an extensive benchmark across four datasets and two definitions of OoD. Our experiments show that the robustness of all metrics under test does not solely depend on the nature of the IoD dataset or the OoD definition, but also on the architecture of the classifier, which stresses the need for thorough experimentations for future work in OoD detection.

Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.

No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics

Submission Guidelines: Yes

Please Choose The Closest Area That Your Submission Falls Into: Deep Learning and representational learning