News

Multimodal Perception and Sensing for Autonomous Vehicles

Within the framework of the Ecological and Digital Transition call, the CD6 has been awarded the USEFUL project (TED2021-132338B-I00). Under the title "Multimodal Perception and Sensing for Autonomous Vehicles", the project will carry out the complete construction of an instrumented vehicle for testing different sensor configurations (cameras at different wavelengths, 3D lidar, thermal imaging). These sensors will be used to capture and generate a dataset of annotated data of high multimodality. In an advanced computing cluster developed for the project, these datasets will be used to improve existing Deep Learning algorithms for their application in the perception of vulnerable vehicles (pedestrians, cyclists, microvehicles, ...). The objective is to build better ADAS (Advanced Driver Assistance Systems) sensors that increase the safety of all road users. This will reduce the accident rate of vehicles, moving towards autonomous vehicles with better algorithms and sensors, which will contribute to both the ecological and digital transition of different mobility means. Finally, the vehicle will end up representing a small support infrastructure for companies and researchers interested in integrating their sensors and testing them on a real vehicle.

Image source: nuscenes.org