Neural networks to generate an image from a viewpoint

A system, comprising:

• one or more processors to:
• use one or more first neural networks to generate depth information of an environment around an object based, at least in part, on images of the environment captured by cameras at different locations on the object;
• use one or more second neural networks to generate feature information based, at least in part, on the images;
• generate, based, at least in part, on the depth information and the feature information, a data structure including points in 3-dimensional (3D) space with feature and depth values for respective ones of the points;
• and
• use one or more third neural networks to generate one or more first images from a viewpoint outside of the object based, at least in part, on the data structure.

A method, comprising:

• using one or more first neural networks to generate depth information of an environment around an object based, at least in part, on images of the environment captured by cameras at different locations on the object;
• using one or more second neural networks to generate feature information based, at least in part, on the images;
• generating, based, at least in part, on the depth information and the feature information, a data structure including points in 3-dimensional (3D) space with feature and depth values for respective ones of the points;
• and
• using one or more third neural networks to generate one or more first images from a viewpoint outside of the object based, at least in part, on the data structure.