Systems and methods for fully coupled models for crowd navigation

A method for crowd navigation of a host vehicle among a plurality of agents, comprising:

• identifying a number of agents in a physical environment based on sensor data from a host;
• calculating a set of Gaussian processes, wherein the set of Gaussian processes includes a Gaussian Process for each agent of the number of agents;
• determining an objective function based on an intent and a flexibility for the host and at least two agents of the plurality of agents to classify first order interactions and second order interactions, wherein the first order interactions include interactions between the host and one of the at least two agents, and wherein the second order interactions include interactions between agents of the at least two agents;
• and
• generating a model of the number of agents by applying the objective function to the set of Gaussian processes, wherein the model includes a convex configuration of the physical environment.

A non-transitory computer readable storage medium storing instructions that when executed by a computer having a processor to perform a method for crowd navigation of a host vehicle among a plurality of agents, the method comprising:

• identifying a number of agents in a physical environment based on sensor data received from a host;
• calculating a set of Gaussian processes, wherein the set of Gaussian processes includes a Gaussian Process for each agent of the number of agents;
• determining an objective function based on an intent and a flexibility for the host and at least two agents of the plurality of agents to classify first order interactions and second order interactions, wherein the first order interactions include interactions between the host and one of the at least two agents, and wherein the second order interactions include interactions between agents of the at least two agents;
• and
• generating a model of the number of agents by applying the objective function to the set of Gaussian processes, wherein the model includes a convex configuration of the physical environment.