General Bitwidth Assignment for Efficient Deep Convolutional Neural Network Quantization | IEEE Journals & Magazine | IEEE Xplore