Contextualised Out-of-Distribution Detection using Pattern Identication
This work addresses robust OoD detection for visual classifiers, providing a method that is OoD-agnostic and includes a new benchmark, but it is incremental as it builds on existing explainable AI techniques.
The authors tackled out-of-distribution detection for visual classifiers by proposing CODE, which uses class-specific pattern identification without retraining, and introduced a new benchmark with perturbations to quantify dataset discrepancies, achieving competitive results on standard benchmarks.
In this work, we propose CODE, an extension of existing work from the field of explainable AI that identifies class-specific recurring patterns to build a robust Out-of-Distribution (OoD) detection method for visual classifiers. CODE does not require any classifier retraining and is OoD-agnostic, i.e., tuned directly to the training dataset. Crucially, pattern identification allows us to provide images from the In-Distribution (ID) dataset as reference data to provide additional context to the confidence scores. In addition, we introduce a new benchmark based on perturbations of the ID dataset that provides a known and quantifiable measure of the discrepancy between the ID and OoD datasets serving as a reference value for the comparison between OoD detection methods.