System and method for vehicle lane change prediction using structural recurrent neural networks

A non-transitory computer-readable medium storing instructions for predicting lane changes for nearby vehicles of a host vehicle and that, when executed by one or more processors, cause the one or more processors to:

• in response to detecting that one or more of the nearby vehicles are present nearby the host vehicle, collect, using at least one sensor of the host vehicle, pose information about the nearby vehicles, the pose information including at least one of a position, an orientation, and a trajectory of the one or more of the nearby vehicles relative to the host vehicle,
• wherein the one or more of the nearby vehicles are traveling in lanes of a roadway proximate to the host vehicle and in a direction of travel of the host vehicle;
• analyze the pose information of the nearby vehicles using separate recurrent units of a structural recurrent neural network (S-RNN) to generate factors according to the lanes,
• wherein the instructions to analyze the pose information include instructions to associate the nearby vehicles with a discrete set of separate positions defined by a factor graph relative to the host vehicle,
• wherein the recurrent units correlate with factors of respective ones of the lanes and spatiotemporal interactions of respective ones of the nearby vehicles in the respective ones of the lanes, and wherein the recurrent units are mapped to the discrete set of separate positions from the factor graph including separate ones of the recurrent units for separate lanes about the host vehicle;
• generate prediction indicators for the nearby vehicles as a function of the factors for the lanes using the S-RNN;
• and
• provide electronic outputs identifying the prediction indicators that specify a likelihood of the nearby vehicles changing between the lanes.

A method of predicting lane changes for nearby vehicles of a host vehicle, comprising:

• in response to detecting that one or more of the nearby vehicles are present nearby the host vehicle, collecting, using at least one sensor of the host vehicle, pose information about the nearby vehicles, the pose information including at least one of a position, an orientation, and a trajectory of the one or more of the nearby vehicles relative to the host vehicle,
• wherein the one or more nearby vehicles are traveling in lanes of a roadway proximate to the host vehicle and in a direction of travel of the host vehicle;
• analyzing the pose information of the nearby vehicles using separate recurrent units of a structural recurrent neural network (S-RNN) to generate factors according to the lanes,
• wherein analyzing the pose information includes associating the nearby vehicles with a discrete set of separate positions defined by a factor graph relative to the host vehicle,
• wherein the recurrent units correlate with factors of respective ones of the lanes and spatiotemporal interactions of respective ones of the nearby vehicles in the respective ones of the lanes, and wherein the recurrent units are mapped to the discrete set of separate positions from the factor graph including separate ones of the recurrent units for separate lanes about the host vehicle;
• generating prediction indicators for the nearby vehicles as a function of the factors for the lanes using the S-RNN;
• and
• providing electronic outputs identifying the prediction indicators that specify a likelihood of the nearby vehicles changing between the lanes.