Teach driverless vehicles to predict pedestrian movement

By focusing on the gait, body symmetry and foot placement of humans, University of Michigan (U-M) researchers are teaching self-driving cars to predict pedestrian movements with greater accuracy..

Data collected by vehicles through cameras, LiDAR and GPS allow researchers to capture video clips of humans in motion and then recreate them in 3D computer simulation. With that, they’ve created a “biomechanically inspired recurrent neural network” that catalogues human movements.

According to the researchers, they can use this to predict poses and future locations for one or several pedestrians up to about 50 yards from the vehicle.