FPGA-Based Reduction Techniques for Efficient Deep Neural Network Deployment | IEEE Conference Publication | IEEE Xplore