基于人工神经网络说话人识别的身份认证方案

摘要

Due to the CPSO--BP neural network introducing chaotic idea in Particle Swarm Optimization, which improvsing the ability of particle swarm optimization algorithm to get rid of the local extreme points and enhances the accuracy and convergence speed of the BP network, the recognition rate and speed of training in the speaker recognition based on BP neural network have been improved greatly. In this paper, we used the identification technique to the identity authentication, and proposed a new biometric--based authentication scheme. The performance analysis results shows that the scheme is safe and effective.%由于CPSO--BP神经网络通过在粒子群优化算法中引入混沌思想，改善了粒子群优化算法摆脱局部极值点的能力，提高了BP网络的计算精度和收敛速度，使得基于BP神经网络的说话人识别其识别率和训练速度都得到较大提高。本文将该识别技术应用到身份认证中，提出了一种新的基于生物特征的身份认证方案，通过对其性能的分析研究，新的方案是安全有效的。