Why Normalizing Flows Fail to Detect Out-of-Distribution Data
Detecting out-of-distribution (OOD) data is crucial for robust machine learning systems. Normalizing flows are flexible deep generative models that often surprisingly fail to distinguish between in- and out-of-distribution data: a flow trained on pictures of clothing assigns higher likelihood to handwritten digits. We investigate why normalizing flows perform poorly for OOD detection. We demonstrate that flows learn local pixel correlations and generic image-to-latent-space transformations which are not specific to the target image dataset. We show that by modifying the architecture of flow coupling layers we can bias the flow towards learning the semantic structure of the target data, improving OOD detection. Our investigation reveals that properties that enable flows to generate high-fidelity images can have a detrimental effect on OOD detection.
Code (1)
Tasks
Out of Distribution (OOD) DetectionMethods 이 논문이 사용한 방법론
Similar Papers 제목 키워드 기반
InFlow: Robust outlier detection utilizing Normalizing Flows
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) inp…
Density EstimationOutlier DetectionOut of Distribution (OOD) DetectionOut-of-distribution detection using normalizing flows on the data manifold
Using the intuition that out-of-distribution data have lower likelihoods, a common approach for out-of-distribution detection involves estimating the underlying data distribution. Normalizing flows are likelihood-based g…
Density EstimationOut-of-Distribution DetectionFANFOLD: Graph Normalizing Flows-driven Asymmetric Network for Unsupervised Graph-Level Anomaly Detection
Unsupervised graph-level anomaly detection (UGAD) has attracted increasing interest due to its widespread application. In recent studies, knowledge distillation-based methods have been widely used in unsupervised anomaly…
Anomaly DetectionKnowledge DistillationUnsupervised Anomaly DetectionSliced-Wasserstein normalizing flows: beyond maximum likelihood training
Despite their advantages, normalizing flows generally suffer from several shortcomings including their tendency to generate unrealistic data (e.g., images) and their failing to detect out-of-distribution data. One reason…
Feature Density Estimation for Out-of-Distribution Detection via Normalizing Flows
Out-of-distribution (OOD) detection is a critical task for safe deployment of learning systems in the open world setting. In this work, we investigate the use of feature density estimation via normalizing flows for OOD d…
Density Estimationimage-classificationImage ClassificationOut-of-Distribution Detection+1