Information on Collective Perception in Autonomous Driving

Content

Vehicle-to-everything (V2X) communication promises to improve the safety and comfort of autonomous vehicles. Collective perception, naturally extends the V2X paradigm, by having multiple vehicles contributing to the the perception of the environment. This approach yields the advantage that also non-connected objects and vehicles can be perceived, decreasing the required minimal market penetration to make the approach efficient. In turn this can increase the incentive for the automotive industry to adapt the approach. In this module we focus on state-of-the-art collective perception research and cover interdisciplinary fundamental topics as wireless networking, computer vision, data fusion and sensors.

Pre-requisits

None. The lecture “Cooperative Autonomous Vehicles” is a good addition and helpful to grasp the content of this lecture but not necessary.

Qualification objectives

Students

• know the fundamentals of vehicular communications and networking
• look critically into current research topics in the field of autonomous driving,
• learn basic concepts in computer vision, perception and machine learning,
• apply methods to fuse multi-modal data and employ collaborative algorithms,
• use simulation tools to model autonomous vehicles.

Exercise

The lecture is accompanied with an exercise.

Exam

Written exam. No multiple-choice.