基于说话人聚类和支持向量机的说话人确认研究

说话人确认系统需要用目标说话人和背景模型说话人的语音数据对模型进行训练。背景模型说话人可随机选取或选取与目标说话人相近的说话人。采用说话人聚类的方法可以有效地解决说话人背景模型的选取问题。支持向量机用作说话人确认模型来训练目标说话人和背景说话人的语音数据。实验表明该方法对与文本无关的说话人确认问题是有效的。

Speaker Verification Based on Speaker Clustering and Support Vector Machines

In conventional speaker verification, to construct good speaker background models is difficult. A novel approach to speaker verification is proposed: the background speakers are clustered into groups, and the background model is constructed based on those groups. The support vector machine based speaker verification models are trained on the enrolled speaker and the background model. Preliminary results on text independent speaker verification are provided to demonstrate the effectiveness of such systems.