"Cyber Researchers at Ben-Gurion University Fool Autonomous Vehicle Systems With Phantom Images"
Researchers at Ben-Gurion University of the Negev's (BGU) Cyber Security Research Center projected "phantom" images on the road and a digital billboard to fool the autopilot of an autonomous vehicle into applying its brakes, following fake traffic signs, and driving into the opposite lane. A new research paper titled, "Phantom of the ADAS," brings attention to the perceptual challenge faced by autopilots and advanced driving assistance systems (ADASs) in semi-autonomous or fully autonomous cars and the potential exploitation of this challenge by attackers. The BGU researchers demonstrated that these autonomous vehicle systems could mistake depthless projections of objects (phantom) for real objects. This perceptual problem can be abused by attackers to manipulate the autonomous vehicle, posing a significant threat to the safety of drivers and passengers. According to the researchers, this type of attack could be performed using a commercial drone and an inexpensive image projector. This article continues to discuss the demonstrated uses of phantom images to trick autonomous vehicle systems and the development of a neural network model to detect phantoms with high accuracy.