TY - RPRT TI - Enhancing speaker identification performance under the shouted talking condition using second-order circular hidden Markov models AU - Ismail Shahin PY - 2017 DO - 10.1016/j.specom.2006.01.005 UR - https://arxiv.org/abs/1706.09716 ID - 1706.09716 ER -