A Survey on Assessing the Generalization Envelope of Deep Neural Networks: Predictive Uncertainty, Out-of-distribution and Adversarial Samples
It addresses the practical need for reliable DNN deployment by synthesizing fragmented literature, but it is incremental as it surveys and connects existing approaches rather than introducing new methods.
This survey tackles the problem of assessing when deep neural networks (DNNs) generalize correctly by connecting three independent research fields—predictive uncertainty, out-of-distribution detection, and adversarial example detection—into a unified framework for evaluating the generalization envelope of DNNs, providing a structured overview of methods to determine if inputs are within this envelope at inference time.
Deep Neural Networks (DNNs) achieve state-of-the-art performance on numerous applications. However, it is difficult to tell beforehand if a DNN receiving an input will deliver the correct output since their decision criteria are usually nontransparent. A DNN delivers the correct output if the input is within the area enclosed by its generalization envelope. In this case, the information contained in the input sample is processed reasonably by the network. It is of large practical importance to assess at inference time if a DNN generalizes correctly. Currently, the approaches to achieve this goal are investigated in different problem set-ups rather independently from one another, leading to three main research and literature fields: predictive uncertainty, out-of-distribution detection and adversarial example detection. This survey connects the three fields within the larger framework of investigating the generalization performance of machine learning methods and in particular DNNs. We underline the common ground, point at the most promising approaches and give a structured overview of the methods that provide at inference time means to establish if the current input is within the generalization envelope of a DNN.