Efficient deep neural network acceleration through FPGA-based batch processing | IEEE Conference Publication | IEEE Xplore