Improving the Post-Training Neural Network Quantization by Prepositive Feature Quantization | IEEE Journals & Magazine | IEEE Xplore