InFlow: Robust outlier detection utilizing Normalizing Flows
This addresses a robustness problem in deep generative models for researchers and practitioners in machine learning, offering an incremental improvement over existing methods.
The paper tackles the overconfidence issue of normalizing flows in outlier detection by extending them with an attention mechanism, achieving state-of-the-art performance without requiring outlier data for training.
Normalizing flows are prominent deep generative models that provide tractable probability distributions and efficient density estimation. However, they are well known to fail while detecting Out-of-Distribution (OOD) inputs as they directly encode the local features of the input representations in their latent space. In this paper, we solve this overconfidence issue of normalizing flows by demonstrating that flows, if extended by an attention mechanism, can reliably detect outliers including adversarial attacks. Our approach does not require outlier data for training and we showcase the efficiency of our method for OOD detection by reporting state-of-the-art performance in diverse experimental settings. Code available at https://github.com/ComputationalRadiationPhysics/InFlow .