Out-of-Distribution Detection with Prototypical Outlier Proxy
This work addresses the critical issue of over-confidence in deep models for real-world deployment, offering a more efficient and effective solution for out-of-distribution detection.
The paper tackles the problem of out-of-distribution detection in deep learning by proposing the Prototypical Outlier Proxy framework, which reduces average FPR95 by 7.70%, 6.30%, and 5.42% on benchmarks like CIFAR-10, CIFAR-100, and ImageNet-200 compared to previous methods, while also training 7.2X faster and inferring 19.5X faster than a recent approach.
Out-of-distribution (OOD) detection is a crucial task for deploying deep learning models in the wild. One of the major challenges is that well-trained deep models tend to perform over-confidence on unseen test data. Recent research attempts to leverage real or synthetic outliers to mitigate the issue, which may significantly increase computational costs and be biased toward specific outlier characteristics. In this paper, we propose a simple yet effective framework, Prototypical Outlier Proxy (POP), which introduces virtual OOD prototypes to reshape the decision boundaries between ID and OOD data. Specifically, we transform the learnable classifier into a fixed one and augment it with a set of prototypical weight vectors. Then, we introduce a hierarchical similarity boundary loss to impose adaptive penalties depending on the degree of misclassification. Extensive experiments across various benchmarks demonstrate the effectiveness of POP. Notably, POP achieves average FPR95 reductions of 7.70%, 6.30%, and 5.42% over the second-best methods on CIFAR-10, CIFAR-100, and ImageNet-200, respectively. Moreover, compared to the recent method NPOS, which relies on outlier synthesis, POP trains 7.2X faster and performs inference 19.5X faster. The source code is available at: https://github.com/gmr523/pop.