Urban Driving

Urban driving research focuses on developing safe and efficient autonomous navigation systems capable of handling complex, dynamic environments. Current efforts concentrate on improving trajectory planning algorithms, often employing hierarchical reinforcement learning, model predictive control, and generative models to predict and react to other road users' behavior, leveraging representations like velocity fields and risk maps. These advancements aim to enhance the reliability and robustness of autonomous vehicles, addressing challenges such as occlusion, adverse weather conditions, and the need for interpretable and verifiable decision-making processes. Ultimately, this research contributes to safer and more efficient urban transportation systems.