An adaptive comb filtering method as applied to acoustic analyses of pathological voice
An adaptive comb filtering method as applied to acoustic analyses of pathological voice
复制标题
一种应用于病态语音声学分析的自适应梳状滤波方法
DOI：
10.1109/icassp.1986.1168996
复制
发表时间：
1986
期刊：
影响因子：
--
通讯作者：
Y. Kikuchi
中科院分区：
文献类型：
--
作者：
H. Kasuya;S. Ogawa;Y. Kikuchi
一种自适应梳状滤波方法，最初由Frazier等人研究。[5]为了增强由于加性噪声引起的退化语音，应用于估计由于声带的病理变化而在声门处产生的声音噪声。在应用该方法时，特别强调了音调周期持续时间的准确确定。通过一种新的声学度量-归一化噪声能量(NNE)来量化估计的嗓音噪声量。用正常对照组和50例声门癌患者的元音样本进行的实验表明，NNE对区分正常嗓音和病理嗓音状态是有用的。
An adaptive comb filtering method, which was initially investigated by Frazier et al. [5] for the enhancement of degraded speech due to additive noise, is applied for estimating vocal noise generated at the glottis due to pathological changes of the vocal folds. In applying the method, special emphasis is placed on the accurate determination of pitch period durations. The amount of estimated vocal noise is quantified by a novel acoustic measure, normalized noise energy(NNE). Experiments with vowel samples spoken by 64 normal control subjects and 50 patients with glottic cancer have shown that the NNE is useful for the distinction between normal and pathological voice status.