Efficient Neural Network Compression through Booth Coding and Exponential of Two Quantization for Enhanced Inference Performance | IEEE Conference Publication | IEEE Xplore