Neural network-based image compression

A processor, comprising:

• one or more circuits to:
• generate first residual information determined in a pixel domain for one or more images based, at least in part, on one or more other images;
• use one or more neural networks to generate image enhancement data for the one or more images including second residual information determined in a neural network feature space;
• and
• generate an encoding of the one or more images based, at least in part, upon the first residual information determined in the pixel domain and the second residual information determined in the neural network feature space.