Intelligent Vehicles: Communication, Sensing and Perception
More information: https://static.studiegids.tue.nl
In this mini-program, students will be introduced to the fundamental communication, sensing and perception technologies for intelligent vehicles. It consists of 3 courses given by experts from TU/e and from the automotive industry. These courses include lab experiments, to give students hands-on experience with state-of-the-art technology. Students following this mini program, furthermore have the opportunity to do their final bachelor project in this context, so as to obtain real-world immersion in the subject matter.
Automotive Sensing (5XSJ0)
More information: http://oase.tue.nl
Sensing the environment around the vehicle is one of the most important aspects of autonomous driving. In this course, we introduce the most important sensor modalities, i.e. RADAR, LIDAR, thermal and stereo vision, active 3D vision, GPS(-INS), and ultrasound. For each sensor modality the students are taught: (1) the use of the sensor in automotive applications, (2) the physical principles underlying the sensor, (3) the mathematical models describing these physical principles, (4) how to derive first-order probabilistic models describing the accuracy of the sensor, (5) the fundamental pro's and con's of the sensor, (6) the recent advances in improving the sensor. The lectures will be given by TUE researchers as well as by experts from industry. The course includes extensive lab experiments with real sensors.
Data fusion & Semantic interpretation (5XSK0)
More information: http://oase.tue.nl
Sensing the environment around the vehicle requires giving real-world significance to sensory signals. We call this “semantic interpretation”. For example, the vehicle must be able to decide whether a pixel in an image belongs to a tree or to a pedestrian. This semantic interpretation is done by advanced pattern recognition software, of which the principles are taught in this course. For intelligent vehicles to be safe, the interpretation of sensory signals must be done extremely reliably in a wide variety of environmental conditions. This is achieved by fusing signals from multiple and, most importantly, from different sensor modalities, as this allows mitigating the pro's and con's of different sensors. Sensor fusion is performed by probabilistic filtering techniques, of which the most well-known is the Kalman filter. This and other more advanced filtering techniques, such as particle filtering, will be taught in this course. Besides lectures, the course also incorporates programming assignments in which students will develop a pedestrian detection system.
