声纹识别在校园网身份认证中的作用

说话人识别是语音识别的一种特殊方式,对于各种领域的身份认证具有得天独厚的优势.以线性预测系数（Linear Prediction Coefficients,简称LPC）作为特征参数,采用动态时间规整（DTW）算法建立了一个与文本相关说话人确认系统,它基于C/S（客户/服务器）模型,采用TCP/IP,能够实现局域网上的语音登陆.实验结果表明,这种系统在通过校园网进行说话人身份认证中是有效的.

Research and realization of the campus net ID authentication system

Speaker recognition is a special way of speech recognition, which has many advantages in identity verification. In this paper, a text - dependent speaker verification system is set up based on DTW. Linear Prediction Coefficients （LPC） are extracted from speech signal and formed feature vectors. These features are transported to a recognition server through the campus net using TCP/IP. Experimental results indicate that this system is effective for speaker verification.