Method and system for behavioral cloning of autonomous driving policies for safe autonomous agents

A non-transitory computer-readable medium having program code recorded thereon for behavior cloned vehicle trajectory planning, the program code being executed by a processor and comprising:

• program code to perceive vehicles proximate an ego vehicle in a driving environment, including a scalar confidence value of each perceived vehicle;
• program code to generate a bird's-eye-view (BEV) grid showing the ego vehicle and each perceived vehicle based on each of the scalar confidence values by program code to compute a 3D tensor ϕ, where each dimension of the 3D tensor ϕ represents a category of an estimated state together with a respective estimated confidence value;
• program code to ignore at least one of the perceived vehicles when the scalar confidence value of the at least one of the perceived vehicles is less than a predetermined value;
• and
• program code to select an ego vehicle trajectory based on a cloned expert vehicle behavior policy according to remaining perceived vehicles.

A system for behavior cloned vehicle trajectory planning, the system comprising:

• a vehicle perception module configured to perceive vehicles proximate an ego vehicle in a driving environment, including a scalar confidence value of each perceived vehicle;
• a semantic grid module configured to generate a bird's-eye-view (BEV) grid showing the ego vehicle and each perceived vehicle based on the scalar confidence values by computing a 3D tensor ϕ, where each dimension of the 3D tensor ϕ represents a category of an estimated state together with a respective estimated confidence value;
• a behavior cloning model configured to ignore at least one of the perceived vehicles when the scalar confidence value of the at least one of the perceived vehicles is less than a predetermined value;
• and
• a vehicle trajectory selection module configured to select an ego vehicle trajectory based on a cloned expert vehicle behavior policy according to remaining perceived vehicles.