Drone Racing

Visual attention prediction improves performance of autonomous drone racing agents

We leverage human eye-gaze data to train a deep learning controller for vision-based autonomous drone racing. Our controller achieves performances as good as human pilots.

Expertise affects drone racing performance

We compare drone racing performances among human pilots in real-world racing. Our analysis shows that experts consistently achieve more optimal racing lines, maneuver execution, and faster laptimes than beginners.

Human-piloted drone racing: Visual processing and control

We This work investigates the relationship between human eye movements, control behavior, and flight performance in a drone racing task.