HYBOOD: A Hybrid Generative Model for Out-of-Distribution Detection with Corruption Estimation

We propose HYBOOD, a hybrid out-of-distribution model based on normalizing flow followed by a simple linear classification model. In real-world settings, it is known that data corruption has a strong influence on model degradation; for example image quality like noise, blur and image geometry like translation, scaling, rotation. MNIST-C, CIFAR10-C are the general synthesized datasets to measure model robustness and corruption difficulty in terms of covariate and semantic shifts. HYBOOD shows that the separability between indistribution, covariate shift, and semantic shift can be represented by generative distribution distance and log-scale density. We also find out the types of covariate shifts are ordered by corruption difficulty ranking (CDR) for the datasets. To the best of our knowledge, this is the first method to measure data corruption difficulty with generative models using Wasserstein Distance, Mutual Information and Minimal Description Length. In this paper, we pose interesting experimental results that the generative model tested on MNIST-C is most deteriorated by fog, impulse noise and stripe corruption types. This can be interpreted that those types are challenging corruptions to the generative model in uncertainty and complexity. By training in-distribution data only, HYBOOD achieves out-of-distribution detection performance for distinguishable covariate and semantic shifts, and quantifying covariate shift ranking.