Abstract: An image predictor is trained to produce a predicted image based on N preceding images captured by a vehicle camera and vehicle controls. A discriminator is trained to distinguish between an image following P preceding images in an image stream and one that is not a subsequent image. A control generator generates estimated controls based on a set of N images and the estimated controls and set of N images are input to the image predictor. A predicted image and the set of N images are input to the image predictor which outputs a value indicating whether the predicted image is accurate. A loss function based on this value and a difference between the vehicle controls and the estimated controls for the set of N images is used as feedback for training the control generator.

Abstract: A high-level vehicle command is determined based on a location of the vehicle with respect to a route including a start location and a finish location. An image is acquired of the vehicle external environment. Steering, braking, and powertrain commands are determined based on inputting the high-level command and the image into a Deep Neural Network. The vehicle is operated by actuating vehicle components based on the steering, braking and powertrain commands.

Abstract: A high-level vehicle command is determined based on a location of the vehicle with respect to a route including a start location and a finish location. An image is acquired of the vehicle external environment. Steering, braking, and powertrain commands are determined based on inputting the high-level command and the image into a Deep Neural Network. The vehicle is operated by actuating vehicle components based on the steering, braking and powertrain commands.

Abstract: An image predictor is trained to produce a predicted image based on N preceding images captured by a vehicle camera and vehicle controls. A discriminator is trained to distinguish between an image following P preceding images in an image stream and one that is not a subsequent image. A control generator generates estimated controls based on a set of N images and the estimated controls and set of N images are input to the image predictor. A predicted image and the set of N images are input to the image predictor which outputs a value indicating whether the predicted image is accurate. A loss function based on this value and a difference between the vehicle controls and the estimated controls for the set of N images is used as feedback for training the control generator.