Join us to hear about the challenges of obtaining accurate real-life data for edge case evaluation and how synthetic data can fill this gap. We'll highlight the relevant camera edge cases when simulated with a synthetic data approach.
Venue:
Virtual
Training and evaluating perception systems requires a large variety of accurate & annotated data to ensure the safety of Autonomous Vehicles (AV) in their Operational Design Domain (ODD). Acquiring enough real-life data for sensor perception edge cases or dangerous (critical) driving situations is challenging. Synthetic data can fill this gap. However, a physics-based simulation approach is necessary to obtain trustable synthetic data.
This synthetic data is used for multispectral light and electromagnetic wave propagation in 3D synthetic scenes and sensor modeling. In this webinar, we intend to highlight the relevant camera edge cases when simulated with this approach and the necessity of accurate synthetic data to solve this problem.
All AVxcelerate users, ADAS & AD engineers, HiL, SiL test engineers should attend to learn why synthetic data is necessary for safety by validation especially while evaluating the edge cases.
Lionel Bennes
