似然比在法庭说话人确认中的应用

本文把成功应用于DNA检验的证据评估方法（似然比）应用于法庭语音证据评估之中,提取语音的LPC作为识别特征,并使用45人电话对话录音中元音/a/作为样本进行了测试。结果表明该方法不仅能正确识别说话人,而且能根据当前嫌疑人样本和问题语音样本的差异,量化该语音样本作为证据的力度,为法庭提供科学合理的证据评估结果和科学解释。同时,自动特征提取的引入比起人工提取共振峰特征,提高了工作效率,识别系统性能也获得大幅提升。

The Application of Likelihood Ratio in Forensic Speaker Verification.

A forensic speaker identification method based on likelihood ratio and LPC features is presented in this paper.Likelihood ratio has been successfully applied in the evaluation of DNA,but it is a new application in forensic speaker verification context.This method is tested in vowel /a/ of 45 people’s telephone dialog recordings and shows high identification ratio.Not only can the method identify the speaker,but also quantify the evidence strength according the acoustic difference between the questioned sample and the suspect’s sample,and provide the court with the scientific and reasonable evaluation result.Compared to the manual method of formants extraction,the usage of auto extraction of features increases the efficiency and performance of the identification.