Emil Barr

Emil Barr, a fresh PhD graduate and researcher at FAIR, dives into the fascinating world of navigation in robotics. He discusses the shift from traditional mathematical models to cutting-edge neural networks that help robots navigate dynamic environments. Topics include the complexities of training world models with observational data, the innovative use of conditional diffusion transformers for state prediction, and the role of advanced trajectory generation methods. It's a thrilling exploration of how AI is shaping the future of robotics!