说话人听觉识别研究综述  

Research on the Speaker's Audio Identity

作　　者：杨俊杰[1] 李剑[2] 

机构地区：[1]山西警官高等专科学校,山西太原030021 [2]贵州警官职业学院,贵州贵阳550005

出　　处：《贵州警官职业学院学报》2010年第6期64-67,共4页Journal of Guizhou Police Officer Vocational College

基　　金：贵州省科技厅2008年科技支撑项目(黔科合SY[2008]3064号)

摘　　要：说话人的听觉识别一直都是司法话者识别的重要方法之一。国外实验表明:听觉识别率随时间的延长在降低,它与样本的长度、质量、内容以及听辨人的身份相关而与听辨人的性别无关。国外,亲听证人听辨法已经成为一种常见的司法话者识别方法,而我国实践中却应用得较少,因此有必要加强相关理论的应用研究。The speaker＇s audio identity is always one of the most important ways of forensic speaker recognition.The foreign experiment shows： the ratio of audio identity is lowering with the prolongation of time,and it has something with the length,quality and content of sample and listener＇s identity,not with the listener＇s gender.In foreign countries,the earwitness＇way of listening and information receiving has become a kind of way of forensic speaker recognition,but seldom used in practices in China,so it is necessary to enhance the applied research of relative theory.

关 键 词：司法话者识别 听觉识别 亲听证人听辨法 

分 类 号：D918.2[政治法律—法学]