基于Teager能量算子(TEO)基频的应力影响下的变异语音分类  

TEO-Pitch based classification of stressed speech under G-Force

作　　者：马永林[1] 韩纪庆[1] 张磊[1] 吕成国[1] 王承发[1] 

机构地区：[1]哈尔滨工业大学计算机科学与工程系,哈尔滨150001

出　　处：《声学学报》2002年第6期518-522,共5页Acta Acustica

基　　金：国家自然科学基金(60085001);哈尔滨工业大学跨学科交叉性研究基金(HIT.MD200001)资助项目

摘　　要：变异语音识别是一项极具挑战意义的研究课题,一种解决方法是在前端对语音进行变异分类,然后根据不同变异情况采用相关的处理算法。在各种语音变异中,说话人在战斗机、航天飞机座舱等环境中,身体受到应力（重力）影响时的情况更具有特殊性。其所引起的发音变异有别于因心理的、感知的或生理的因素所引起的变异,目前国内外还鲜见有关应力影响不变异语音分类问题的专门研究。木文从对应力影响下的几种基于基频的语音特征的分析出发,提出了对应力影响下的变异语音和正常语音进行分类的方法。对航空模拟飞行器中采集的小词表实验样本,特定人平均分类正确率达到了93.3％,多说话人分类上确率达到了85.8％。Stressed speech recognition is a challenging work. One approach for addressing the degradation is to utilize front-end stress classification to direct a stress dependent recognition algorithm which separately models each speech production domain. Since G-Force has a direct physical impact on human speech production, G-Force induced stress is differ from stresses induced by psychological or perceptual factors which have been widely studied. So far, there is little research about classification of speech under G-Force. This study proposes a new approach based on several pitch features for the classification of G-Force/Normal speech. Stressed speech database under G-Force with two speakers has been collected in an aero-flight simulator, and the speaker dependent and multi-speaker average classification rate are 93.3% and 85.8% respectively.

关 键 词：TEAGER能量算子 基频 应力 变异语音 分类 语音识别 语音特征 

分 类 号：TN912.3[电子电信—通信与信息系统]