Diversified Outlier Exposure for Out-of-Distribution Detection via Informative Extrapolation
This work addresses the challenge of reliable OOD detection for real-world machine learning applications, offering an incremental improvement over existing outlier exposure methods by enhancing outlier diversity.
The paper tackles the problem of out-of-distribution (OOD) detection by proposing Diversified Outlier Exposure (DivOE), a framework that synthesizes more informative outliers from auxiliary data to improve detection performance, achieving state-of-the-art results on benchmarks like CIFAR-10 and ImageNet.
Out-of-distribution (OOD) detection is important for deploying reliable machine learning models on real-world applications. Recent advances in outlier exposure have shown promising results on OOD detection via fine-tuning model with informatively sampled auxiliary outliers. However, previous methods assume that the collected outliers can be sufficiently large and representative to cover the boundary between ID and OOD data, which might be impractical and challenging. In this work, we propose a novel framework, namely, Diversified Outlier Exposure (DivOE), for effective OOD detection via informative extrapolation based on the given auxiliary outliers. Specifically, DivOE introduces a new learning objective, which diversifies the auxiliary distribution by explicitly synthesizing more informative outliers for extrapolation during training. It leverages a multi-step optimization method to generate novel outliers beyond the original ones, which is compatible with many variants of outlier exposure. Extensive experiments and analyses have been conducted to characterize and demonstrate the effectiveness of the proposed DivOE. The code is publicly available at: https://github.com/tmlr-group/DivOE.