Latent Logic is a spin-out from the University of Oxford’s Computer Science department and is commercialising cutting-edge research into novel techniques in Artificial Intelligence and Machine Learning. The first market for Latent Logic is Autonomous Vehicles, which require extensive testing in simulation before they can be placed on the road. For those simulations to be valuable, they need to represent the real world as accurately as possible. Latent Logic is building intelligent ‘virtual humans’ – drivers, cyclists, and pedestrians – to populate these simulation environments to make them more realistic, meaning the situations encountered by the vehicle under test in simulation are similar to those it will encounter on real roads. These A.I. are agents are trained with state-of-the-art Machine Learning methods, using raw traffic footage as ‘demonstration data’ of how humans behave in traffic in the real world. The techniques developed and used by Latent Logic can be applied in any field where autonomous machines need to interact with humans, such as collaborative robotics or immersive, interactive gaming.  

www.latentlogic.com