Speaker Details

Speaker Company

Jerome Leudet

Jerome Leudet has a strong background in AI. After a masters in AI and VR, in early 2000s, he developed AIs in the gaming industry for more than a decade and worked in industrial simulation. With the rise of AI, he had the opportunity to start a PhD and study how synthetic data could be used for training neural networks. He founded AILiveSim on a mission to create rich interactive worlds to train and test algorithms at first for maritime applications. AILiveSim aims to solve the current and future problems with training intelligent systems.

Presentation

Virtual training of an autonomous surface vehicle

Access to the right data is a necessary condition for a successful autonomous development. Today, in the maritime world, data processing is a resource intensive task. Indeed, counting only on real annotated data is limiting, time consuming and expensive. Sometimes the data required does not simply exist, especially in such a complex and large environment as the open sea. This is why, being able to generate synthetic data via simulation with very realistic models and the possibility of configurable parameters to any digital twins including sensors, weather, waves or piers is a game changer.