LGMLOct 12, 2024

The Best of Both Worlds: On the Dilemma of Out-of-distribution Detection

arXiv:2410.11576v1Has Code
Originality Incremental advance
AI Analysis

This addresses model trustworthiness for real-world AI applications by resolving a key dilemma in OOD detection, though it appears incremental as it builds on existing methods.

The paper tackles the problem in out-of-distribution (OOD) detection where state-of-the-art methods sacrifice OOD generalization for detection performance, and proposes a theory-inspired algorithm that achieves superior OOD detection without compromising generalization, as shown in empirical studies on standard benchmarks.

Out-of-distribution (OOD) detection is essential for model trustworthiness which aims to sensitively identify semantic OOD samples and robustly generalize for covariate-shifted OOD samples. However, we discover that the superior OOD detection performance of state-of-the-art methods is achieved by secretly sacrificing the OOD generalization ability. Specifically, the classification accuracy of these models could deteriorate dramatically when they encounter even minor noise. This phenomenon contradicts the goal of model trustworthiness and severely restricts their applicability in real-world scenarios. What is the hidden reason behind such a limitation? In this work, we theoretically demystify the ``\textit{sensitive-robust}'' dilemma that lies in many existing OOD detection methods. Consequently, a theory-inspired algorithm is induced to overcome such a dilemma. By decoupling the uncertainty learning objective from a Bayesian perspective, the conflict between OOD detection and OOD generalization is naturally harmonized and a dual-optimal performance could be expected. Empirical studies show that our method achieves superior performance on standard benchmarks. To our best knowledge, this work is the first principled OOD detection method that achieves state-of-the-art OOD detection performance without compromising OOD generalization ability. Our code is available at \href{https://github.com/QingyangZhang/DUL}{https://github.com/QingyangZhang/DUL}.

Code Implementations1 repo
Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes