Trained Quantization Thresholds for Accurate and Efficient Fixed-Point Inference of Deep Neural Networks

Jain, Sambhav R., Gural, Albert, Wu, Michael · arXiv · 2019