Bayesian OOD detection with aleatoric uncertainty and outlier exposure
arXiv:2102.12959
Abstract
Typical Bayesian approaches to OOD detection use epistemic uncertainty. Surprisingly from the Bayesian perspective, there are a number of methods that successfully use aleatoric uncertainty to detect OOD points (e.g. Hendryks et al. 2018). In addition, it is difficult to use outlier exposure to improve a Bayesian OOD detection model, as it is not clear whether it is possible or desirable to increase posterior (epistemic) uncertainty at outlier points. We show that a generative model of data curation provides a principled account of aleatoric uncertainty for OOD detection. In particular, aleatoric uncertainty signals a specific type of OOD point: one without a well-defined class-label, and our model of data curation gives a likelihood for these points, giving us a mechanism for conditioning on outlier points and thus performing principled Bayesian outlier exposure. Our principled Bayesian approach, combining aleatoric and epistemic uncertainty with outlier exposure performs better than methods using aleatoric or epistemic alone.
References in corpus (17)
- What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?
- A Baseline for Detecting Misclassified and Out-of-Distribution Examples in Neural Networks
- Deep Anomaly Detection with Outlier Exposure
- Learning Confidence for Out-of-Distribution Detection in Neural Networks
- A Simple Baseline for Bayesian Uncertainty in Deep Learning
- WAIC, but Why? Generative Ensembles for Robust Anomaly Detection
- Decomposition of Uncertainty in Bayesian Deep Learning for Efficient and Risk-sensitive Learning
- Cyclical Stochastic Gradient MCMC for Bayesian Deep Learning
- What Are Bayesian Neural Network Posteriors Really Like?
- BatchEnsemble: An Alternative Approach to Efficient Ensemble and Lifelong Learning
- Reducing Network Agnostophobia
- Open Category Detection with PAC Guarantees
- Out-of-distribution Detection in Classifiers via Generation
- Safer Classification by Synthesis
- Bayesian Neural Network Priors Revisited
- Density of States Estimation for Out-of-Distribution Detection
- The Hidden Uncertainty in a Neural Networks Activations