DQI: A Dynamic Quantization Method for Efficient Convolutional Neural Network Inference Accelerators | IEEE Conference Publication | IEEE Xplore