CITIES & CCS Investigator Prof. Muhammad Shafique was invited as one of the speakers at the 2nd Abu Dhabi Smart City Summit on the 23rd of November, 2021.
His talk focused on the Security of AI Functions in Autonomous Vehicles: Artificial Intelligence (AI) technologies such as Machine Learning (ML) have proliferated in a wide range of applications from Smart Cyber-Physical Systems (CPS) and the Internet of Things (IoT). However, despite their outstanding performance, researchers from prominent groups have demonstrated that advanced ML / AI algorithms (particularly the deep neural networks) are embarrassingly easy to fool, and are vulnerable to a wide range of security attacks and safety issues, thereby posing serious issues for safety-critical applications, such as autonomous vehicles. Although ML vulnerabilities in domains such as vision, image, audio are now well-known, little attention has focused on security attacks on the AI features of connected autonomous vehicles. This talk provides an overview of these security issues and specialized techniques for different types of security attacks and defenses focusing on advanced AI algorithms deployed in autonomous vehicles.
