Fetching the paper…

Regularizing Activation Distribution for Training Binarized Deep Networks · Around