Abstract: This paper addresses learning safe output feedback control laws from partial observations of expert demonstrations. We assume that a model of the system ...