The survey covers everything from feature descriptors, regional and global feature metrics, feature learning architectures, deep learning, neuroscience of vision, neural networks, and detailed example architectures to illustrate computer vision hardware and software optimization methods.