Performance Analysis of Out-of-Distribution Detection on Various Trained Neural Networks

Abstract

Several areas have been improved with Deep Learning during the past years. For non-safety related products adoption of AI and ML is not an issue, whereas in safety critical applications, robustness of such approaches is still an issue. A common challenge for Deep Neural Networks (DNN) occur when exposed to out-of-distribution samples that are previously unseen, where DNNs can yield high confidence predictions despite no prior knowledge of the input. In this paper we analyse two supervisors on two well-known DNNs with varied setups of training and find that the outlier detection performance improves with the quality of the training procedure. We analyse the performance of the supervisor after each epoch during the training cycle, to investigate supervisor performance as the accuracy converges. Understanding the relationship between training results and supervisor performance is valuable to improve robustness of the model and indicates where more work has to be done to create generalized models for safety critical applications.

arXiv comment8 pages, 7 figures, presented at SEAA 2019

References in corpus (10)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• Densely Connected Convolutional Networks
• NIPS 2016 Tutorial: Generative Adversarial Networks
• DeepXplore: Automated Whitebox Testing of Deep Learning Systems
• Enhancing The Reliability of Out-of-distribution Image Detection in Neural Networks
• Guiding Deep Learning System Testing using Surprise Adequacy
• DLFuzz: Differential Fuzzing Testing of Deep Learning Systems
• An Analysis of ISO 26262: Using Machine Learning Safely in Automotive Software
• Towards a Framework to Manage Perceptual Uncertainty for Safe Automated Driving
• Towards Structured Evaluation of Deep Neural Network Supervisors