FlowCon: Out-of-Distribution Detection using Flow-Based Contrastive Learning

TL;DR

FlowCon is a novel density-based out-of-distribution detection method that combines normalizing flow with supervised contrastive learning, achieving robust and efficient OOD detection on standard vision datasets.

Contribution

The paper introduces FlowCon, a new approach that integrates flow-based density estimation with contrastive learning for improved OOD detection.

Findings

Outperforms existing methods on CIFAR-10 and CIFAR-100 datasets.

Provides robust detection under various OOD scenarios.

Demonstrates effective density estimation with tractable computation.

Abstract

Identifying Out-of-distribution (OOD) data is becoming increasingly critical as the real-world applications of deep learning methods expand. Post-hoc methods modify softmax scores fine-tuned on outlier data or leverage intermediate feature layers to identify distinctive patterns between In-Distribution (ID) and OOD samples. Other methods focus on employing diverse OOD samples to learn discrepancies between ID and OOD. These techniques, however, are typically dependent on the quality of the outlier samples assumed. Density-based methods explicitly model class-conditioned distributions but this requires long training time or retraining the classifier. To tackle these issues, we introduce \textit{FlowCon}, a new density-based OOD detection technique. Our main innovation lies in efficiently combining the properties of normalizing flow with supervised contrastive learning, ensuring robust representation learning with tractable density estimation. Empirical evaluation shows the enhanced performance of our method across common vision datasets such as CIFAR-10 and CIFAR-100 pretrained on ResNet18 and WideResNet classifiers. We also perform quantitative analysis using likelihood plots and qualitative visualization using UMAP embeddings and demonstrate the robustness of the proposed method under various OOD contexts. Code will be open-sourced post decision.