BootOOD: Self-Supervised Out-of-Distribution Detection via Synthetic Sample Exposure under Neural Collapse

2025-11-17

AI Summary

To address the challenge of detecting semantically similar out-of-distribution (OOD) samples in security-critical scenarios, this paper proposes BootOOD—a fully self-supervised framework. Methodologically, it exploits neural collapse to synthesize pseudo-OOD features solely from in-distribution (ID) data via lightweight transformations, and introduces a lightweight auxiliary head that performs radius-based classification on feature norms, thereby decoupling ID classification from OOD detection—without requiring any real OOD samples or anomaly exposure training. Its core innovation lies in being the first to establish feature norm disparity as a reliable discriminative signal for detecting semantically close OOD samples, synergistically combined with self-supervised pseudo-OOD generation for end-to-end optimization. On CIFAR-10/100 and ImageNet-200, BootOOD achieves OOD detection performance on par with state-of-the-art anomaly exposure methods, while maintaining or even improving ID classification accuracy.

Technology Category

• Computer Vision: Adversarial Attacks & Robustness
• Machine Learning: Unsupervised & Self-Supervised Learning
• Data Mining & Knowledge Management: Anomaly/Outlier Detection

Application Category

• Security and Privacy: Data transparency and provenance
• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Out-of-distribution (OOD) detection is critical for deploying image classifiers in safety-sensitive environments, yet existing detectors often struggle when OOD samples are semantically similar to the in-distribution (ID) classes. We present BootOOD, a fully self-supervised OOD detection framework that bootstraps exclusively from ID data and is explicitly designed to handle semantically challenging OOD samples. BootOOD synthesizes pseudo-OOD features through simple transformations of ID representations and leverages Neural Collapse (NC), where ID features cluster tightly around class means with consistent feature norms. Unlike prior approaches that aim to constrain OOD features into subspaces orthogonal to the collapsed ID means, BootOOD introduces a lightweight auxiliary head that performs radius-based classification on feature norms. This design decouples OOD detection from the primary classifier and imposes a relaxed requirement: OOD samples are learned to have smaller feature norms than ID features, which is easier to satisfy when ID and OOD are semantically close. Experiments on CIFAR-10, CIFAR-100, and ImageNet-200 show that BootOOD outperforms prior post-hoc methods, surpasses training-based methods without outlier exposure, and is competitive with state-of-the-art outlier-exposure approaches while maintaining or improving ID accuracy.

Problem

• Detecting semantically similar out-of-distribution samples for image classifiers
• Developing self-supervised OOD detection using only in-distribution data
• Handling challenging OOD samples through feature norm differentiation

Innovation

• Synthesizes pseudo-OOD features via ID transformations
• Uses radius-based classification on feature norms
• Decouples OOD detection from the primary classifier