On Out-of-distribution Detection with Energy-based Models

2021/07/03 by Sven Elflein, Elflein, Sven, Bertrand Charpentier +5 · 4 citations
Computer Science · #FOS: Computer and information sciences #Machine Learning (cs.LG) #cs.LG

paper · pdf · doi:10.48550/arxiv.2107.08785

Accepted to ICML 2021 Workshop on Uncertainty & Robustness in Deep Learning

arxiv created 2021/07/03 · arxiv updated 2021/07/20

Abstract

Several density estimation methods have shown to fail to detect out-of-distribution (OOD) samples by assigning higher likelihoods to anomalous data. Energy-based models (EBMs) are flexible, unnormalized density models which seem to be able to improve upon this failure mode. In this work, we provide an extensive study investigating OOD detection with EBMs trained with different approaches on tabular and image data and find that EBMs do not provide consistent advantages. We hypothesize that EBMs do not learn semantic features despite their discriminative structure similar to Normalizing Flows. To verify this hypotheses, we show that supervision and architectural restrictions improve the OOD detection of EBMs independent of the training approach.

Cited by

• ALSA: Anchors in Logit Space for Out-of-Distribution Accuracy Estimation
• BOOD: Boundary-based Out-Of-Distribution Data Generation
• A Variational Information Theoretic Approach to Out-of-Distribution Detection
• Score-Based Training for Energy-Based TTS Models

Related

• Energy Correction Model in the Feature Space for Out-of-Distribution Detection
• Master's Thesis: Out-of-distribution Detection with Energy-based Models
• Out-of-Distribution Detection with Class Ratio Estimation
• Revisiting Energy-Based Model for Out-of-Distribution Detection
• Energy-based Out-of-distribution Detection