Accelerating RNN on FPGA with Efficient Conversion of High-Level Designs to RTL | IEEE Conference Publication | IEEE Xplore