A Less Biased Evaluation of Out-of-distribution Sample Detectors

Abstract

In the real world, a learning system could receive an input that is unlike anything it has seen during training. Unfortunately, out-of-distribution samples can lead to unpredictable behaviour. We need to know whether any given input belongs to the population distribution of the training/evaluation data to prevent unpredictable behaviour in deployed systems. A recent surge of interest in this problem has led to the development of sophisticated techniques in the deep learning literature. However, due to the absence of a standard problem definition or an exhaustive evaluation, it is not evident if we can rely on these methods. What makes this problem different from a typical supervised learning setting is that the distribution of outliers used in training may not be the same as the distribution of outliers encountered in the application. Classical approaches that learn inliers vs. outliers with only two datasets can yield optimistic results. We introduce OD-test, a three-dataset evaluation scheme as a more reliable strategy to assess progress on this problem. We present an exhaustive evaluation of a broad set of methods from related areas on image classification tasks. Contrary to the existing results, we show that for realistic applications of high-dimensional images the previous techniques have low accuracy and are not reliable in practice.

arXiv commentto appear in BMVC 2019; v2 is more compact, with more results

References in corpus (13)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• Auto-Encoding Variational Bayes
• SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model size
• Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning Algorithms
• What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?
• On Calibration of Modern Neural Networks
• A Baseline for Detecting Misclassified and Out-of-Distribution Examples in Neural Networks
• A Simple Unified Framework for Detecting Out-of-Distribution Samples and Adversarial Attacks
• Deep Anomaly Detection with Outlier Exposure
• Novelty Detection with GAN
• Safer Classification by Synthesis
• Out-of-Distribution Detection using Multiple Semantic Label Representations
• Online Learning with Abstention

Cited by in corpus (11)

• Can You Trust Your Model's Uncertainty? Evaluating Predictive Uncertainty Under Dataset Shift
• Failing Loudly: An Empirical Study of Methods for Detecting Dataset Shift
• Scalable agent alignment via reward modeling: a research direction
• Detecting Out-of-Distribution Inputs to Deep Generative Models Using Typicality
• On the Validity of Bayesian Neural Networks for Uncertainty Estimation
• On Last-Layer Algorithms for Classification: Decoupling Representation from Uncertainty Estimation
• Bayesian Variational Autoencoders for Unsupervised Out-of-Distribution Detection
• Unsupervised Anomaly Detection with Adversarial Mirrored AutoEncoders
• Anomalous Example Detection in Deep Learning: A Survey
• Out-of-domain Detection for Natural Language Understanding in Dialog Systems
• Class Anchor Clustering: a Loss for Distance-based Open Set Recognition