A Model for Mediating Multi-Modal Human Intent into Safe Maneuvers for UAVs

Direct human interaction with autonomous UAV systems can be enabled through modalities such as speech, gestures, and graphical interfaces. However, interpreting such inputs as directly executable commands introduces safety risks in dynamic environments. Operator requests may conflict with terrain co...