基于神经网络的语音增强系统设计

Design of Speech Enhancement System Based on Neural Network

利用神经网络设计语音信号增强处理系统,在无噪和含噪条件下,提取语音信号的MFCC系数,用于BP神经网络的训练和识别,最终达到语音信号消噪和提高可懂度的目的。自适应神经网络系统具有非线性映射和自学习能力,能够用于噪声信号的非线性建模。它不仅能够获取信号的最佳估计,并且能够克服信号处理中存在的不确定性。仿真结果表明,该自适应噪声抵消器的设计方法,不仅实现简单,而且节省运行时间,语音增强效果很好。

This paper discusses a new method of speech enhancement based on neural networko.MFCC coefficient of the speech signal can be picked up under noise and non-noise condition, trained and recognized by BP neural network. In the end, we Can cancel noise effectively and enhance the intelligibility. Adaptive neural network not only has the ability of nonlinear mapping but also has the ability of self-learning. So it can be used to achieve the linear model of noise, The method not only achieves the optimal estimate but also can overcome uncertainties in signal processing. The simulation result shows that, this speech enhancement system design method, not only realizes simply, but also saves the running time. The speech enhancement effect is very good.