Automatic pathological voice evaluation based on auditory model
Wang Di, Fu Qiang, Yang Lin, Ping Yu, Yan Yonghong, Feng Ji, (1)中国科学院声学研究所中科信利语音实验室，北京 100080; (2)中国科学院物理研究所，北京 100190; (3)中国科学院物理研究所，北京 100190；中国科学院声学研究所中科信利语音实验室，北京 100080; (4)中国人民解放军总医院，北京 100853 · Acta Physica Sinica · 2008
Sustained vowel analysis is an important method in pathological voice evaluation system. In view of the fact that conventional methods generally analyzes the hand-cut samples in FFT frequency domain, an auditory model is proposed as an accurate and robust method for use in pathological vowel analysis. An automatic sample cutting method increases the stability of analysis. Expertiment showed this method is more effective and robust.