binarized neural networks

These networks strive to reduce computational costs by binarizing the weights and activations of the network, that is, constraining them to be either +1 or -1.