MOUNTAIN VIEW, CA
ITL Researcher – Michael Clifford, and co-authors, received the Best Paper Award and Best Demo award at the VehicleSec 2023 workshop.
Michael, and other researchers, have developed the first physical-world attack against connected, autonomous vehicle (CAV) traffic sign recognition systems that is invisible to humans. Toyota does research on security attacks so that we can better prepare to defend against them. Utilizing Infrared Laser Reflection (ILR), they implemented and evaluated an attack that affects CAV cameras, but that humans can not perceive. They formulated the threat model and requirements for an ILR-based sign perception attack. Next, they evaluated attack effectiveness against popular, convolutional neural network (CNN)-based traffic sign recognition systems. They demonstrated a 100% success rate against stop and speed limit signs in their laboratory evaluation. Finally, they discussed the next steps in their research, which include how to defend our vehicles against this sign perception attack.
WIP: Infrared Laser Reflection Attack Against Traffic Sign Recognition Systems