Seongjun Hahm, Akinori Ito, Shozo Makino, Motoyuki Suzuki: A fast speaker adaptation method using aspect model. INTERSPEECH 2008: 1221-1224