Practical Solutions for Machine Learning Safety in Autonomous Vehicles

Abstract

Autonomous vehicles rely on machine learning to solve challenging tasks in perception and motion planning. However, automotive software safety standards have not fully evolved to address the challenges of machine learning safety such as interpretability, verification, and performance limitations. In this paper, we review and organize practical machine learning safety techniques that can complement engineering safety for machine learning based software in autonomous vehicles. Our organization maps safety strategies to state-of-the-art machine learning techniques in order to enhance dependability and safety of machine learning algorithms. We also discuss security limitations and user experience aspects of machine learning components in autonomous vehicles.

Accepted at Safe AI workshop at AAAI 2020

References in corpus (14)

• A Baseline for Detecting Misclassified and Out-of-Distribution Examples in Neural Networks
• ImageNet-trained CNNs are biased towards texture; increasing shape bias improves accuracy and robustness
• Sanity Checks for Saliency Maps
• Using Pre-Training Can Improve Model Robustness and Uncertainty
• Learning Confidence for Out-of-Distribution Detection in Neural Networks
• Multi-Scale Dense Networks for Resource Efficient Image Classification
• Understanding Measures of Uncertainty for Adversarial Example Detection
• Efficient Formal Safety Analysis of Neural Networks
• AI Safety Gridworlds
• An Analysis of ISO 26262: Using Machine Learning Safely in Automotive Software
• Training verified learners with learned verifiers
• Towards Verified Artificial Intelligence
• Evaluating Uncertainty Quantification in End-to-End Autonomous Driving Control
• Predicting Model Failure using Saliency Maps in Autonomous Driving Systems

Cited by in corpus (8)

• Organization of machine learning based product development as per ISO 26262 and ISO/PAS 21448
• A Comparison of Uncertainty Estimation Approaches in Deep Learning Components for Autonomous Vehicle Applications
• A Unified Benchmark for the Unknown Detection Capability of Deep Neural Networks
• Towards Dependable Autonomous Systems Based on Bayesian Deep Learning Components
• Developing Future Human-Centered Smart Cities: Critical Analysis of Smart City Security, Interpretability, and Ethical Challenges
• Towards Audit Requirements for AI-based Systems in Mobility Applications
• Shifting Transformation Learning for Out-of-Distribution Detection
• Exposing Semantic Segmentation Failures via Maximum Discrepancy Competition