摘要
针对传统声纹识别系统存在的易受环境噪声干扰、电磁兼容性差等问题,提出一种基于神经网络的光纤传感系统声纹识别方法。该方法利用光纤传感器采集声学信号,结合深度学习算法实现声纹特征提取和匹配。通过光纤声学信号预处理、深度学习特征提取、自适应声纹匹配等关键技术,有效提高了识别的准确性和健壮性。
A voice print recognition method of optical fiber optic sensing system based on neural network is proposed to address the issues of susceptibility to environmental noise interference and poor electromagnetic compatibility in traditional voiceprint recognition system.This method utilizes fiber optic sensors to collect acoustic signals and combines deep learning algorithms to achieve voiceprint feature extraction and matching.Through key technologies such as fiber optic acoustic signal preprocessing,deep learning feature extraction,and adaptive voiceprint matching,the accuracy and robustness of recognition have been effectively improved.
作者
马鼎山
MA Dingshan(Haishu Branch of Ningbo Municipal Public Security Bureau,Ningbo 315000,China)
出处
《电声技术》
2024年第10期80-82,共3页
Audio Engineering
关键词
声纹识别
光纤传感
深度学习
voiceprint recognition
fiber optic sensing
deep learning
