The inconvenient truth of sensor data in Assisted and Automated Driving
Interlocutors: Professor Valentina Donzella ( University of Warwick)
This talk will describe some of the current technical challenges of Assisted and Automated driving functions, focusing particularly on aspects related to sensors and perception. One of the biggest challenges associated to sensing is the amount of data produced by the environmental perception sensor suite, and the need to transmit these data without latencies and without hindering perception. Our current work demonstrates that high levels of lossy compression can be applied to automotive video camera data without compromising deep neural network based perception and shows how detection can be optimised to cope with compressed data.
It is possible to register (for free) for the event at this link.