Performing speech recognition on multiple parallel files using continuous hidden Markov models on an FPGA | IEEE Conference Publication | IEEE Xplore