强噪声下基于听觉模型的汉语声调提取  

Chinese Tone Extraction in Extremely Noisy Background

作　　者：戴明扬[1] 余凯[1] 徐柏龄[1] 余崇智[1] 

机构地区：[1]南京大学声学研究所近代声学国家重点实验室,江苏南京210093

出　　处：《应用科学学报》2001年第2期121-126,共6页Journal of Applied Sciences

基　　金：国家自然科学基金资助项目 (69872 0 14)

摘　　要：基于人耳听觉模型和汉语语音的短时平稳特性 ,提出一种鲁棒性的汉语普通话声调提取方法 .采用基于人耳听觉模型的相关图来提取语音信号的基频 ,运用无监督的侧抑制神经网络来模拟人耳侧抑制属性进行基频检测 ,为了克服在低信噪比情况下侧抑制神经网络的误判问题 ,引入了相邻语音帧的语音基频的帧间约束 .试验表明 ,该方法在信噪比很低的条件下 ,仍能较准确地识别出目标语音声调 。This paper proposes a robust Chinese tone extraction algorithm based on the human auditory mechanism and short term stationary of Chinese speech. In this method, we use the pooled correlogram based on human auditory model to extract the pitch of speech. An unsupervised lateral inhibitory network is used to get the peak position, which simulates the lateral inhibitory phenomenon in human auditory system. The pitch restriction between successive frames of speech is imposed to get rid of misjudgement in the output of lateral inhibitory network. As shown in the experiments, the method can extract Chinese tone quite well even in rather low SNR cases. It can separate the individual tone clearly as two speakers talk simultaneously.

关 键 词：听觉模型 基音周期 声调提取 侧抑制神经网络 语音基频 语音识别 强噪声 

分 类 号：TN912.34[电子电信—通信与信息系统]