Regularized Training Framework for Combining Pruning and Quantization to Compress Neural Networks | IEEE Conference Publication | IEEE Xplore