Iterative Deployment Exposure for Unsupervised Out-of-Distribution Detection
This addresses the vulnerability of models to OOD samples in real-world deployment scenarios, such as medical diagnosis, but is incremental as it builds on existing U-OOD detection methods.
The paper tackles the problem of unsupervised out-of-distribution (OOD) detection in deep learning models, particularly in medical imaging, by introducing a more realistic setting called Iterative Deployment Exposure (IDE) and proposing a method (CSO) that improves OOD detection performance, showing significant gains over baselines on three medical imaging modalities.
Deep learning models are vulnerable to performance degradation when encountering out-of-distribution (OOD) images, potentially leading to misdiagnoses and compromised patient care. These shortcomings have led to great interest in the field of OOD detection. Existing unsupervised OOD (U-OOD) detection methods typically assume that OOD samples originate from an unconcentrated distribution complementary to the training distribution, neglecting the reality that deployed models passively accumulate task-specific OOD samples over time. To better reflect this real-world scenario, we introduce Iterative Deployment Exposure (IDE), a novel and more realistic setting for U-OOD detection. We propose CSO, a method for IDE that starts from a U-OOD detector that is agnostic to the OOD distribution and slowly refines it during deployment using observed unlabeled data. CSO uses a new U-OOD scoring function that combines the Mahalanobis distance with a nearest-neighbor approach, along with a novel confidence-scaled few-shot OOD detector to effectively learn from limited OOD examples. We validate our approach on a dedicated benchmark, showing that our method greatly improves upon strong baselines on three medical imaging modalities.