ZEBRA: Explaining Rare Cases Through Outlying Interpretable Concepts

Pedro Madeira, André Carreiro, Alex Gaudio, Luís Rosado, Filipe Soares, Asim Smailagic; Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, 2023, pp. 3782-3788

Abstract

Anomaly detection methods can detect outliers, but what are the properties of an outlier? In this paper, we propose ZEBRA, a novel framework for generating explanations of an outlier based on the analysis of feature rarity in an interpretable feature space. The contributions of our work include: (a) a modular model-agnostic framework for explanations of outliers; (b) a statistical explanation method based on a rarity score and weighted aggregation functions; (c) multimodal explanations combining visual, textual, and numeric explanations. ZEBRA simplifies the mapping of low-level features to high-level concepts to generate multimodal and human-readable explanations of outliers.

Related Material

[pdf]
[bibtex]
@InProceedings{Madeira_2023_CVPR, author = {Madeira, Pedro and Carreiro, Andr\'e and Gaudio, Alex and Rosado, Lu{\'\i}s and Soares, Filipe and Smailagic, Asim}, title = {ZEBRA: Explaining Rare Cases Through Outlying Interpretable Concepts}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops}, month = {June}, year = {2023}, pages = {3782-3788} }