Rapid speaker adaptation based on D-code extracted from BLSTM-RNN in LVCSR | IEEE Conference Publication | IEEE Xplore