Autonomous Driving Perception

Catalog description

For widespread adoption, autonomous vehicles need to demonstrate a very high degree of reliability. As such, the focus of this course is a) identifying the key challenges pertaining to autonomous driving perception, and b) potential solutions to those challenges. To do this, the class explores literature in the systems, computer vision, and robotics communities on key topics such as cooperative perception, collaborative planning, multi-modal sensor fusion, end-to-end neural network designs, AI/ML, and the security implications of autonomous driving. In doing so, the course takes a holistic systems approach to enabling safer, more secure, and more reliable autonomous driving through these approaches. A semester-long research project gives hands-on experience with systems for autonomous driving, with class presentations and a term paper that communicate how the project tackles an open problem in autonomous driving.