Autonomous Racing Drone

Autonomous drone racing is a rapidly evolving research area focused on developing algorithms enabling drones to navigate complex courses at high speeds. Current research emphasizes efficient trajectory generation, often employing iterative learning control or deep learning methods like recurrent neural networks and contrastive learning, to optimize speed and maneuverability, sometimes incorporating swarm control. These advancements are improving state estimation techniques, particularly inertial odometry, and enhancing vision-based navigation through methods such as visual attention prediction, leading to more robust and agile autonomous flight capabilities with potential applications in various fields requiring fast and precise aerial maneuvers.