基于深度学习的语音识别系统实现方法

扫码查看

原文链接

NETL
NSTL
万方数据

中文摘要：研究基于深度学习的语音识别系统实现方法,首先探讨语音识别系统的总体框架,其次深入研究梅尔倒谱系数(Mel-Frequency Cepstral Coefficient,MFCC)的提取和深度卷积神经网络(Deep Convolutional Neural Network,DCNN)的基本原理,最后基于Python和PyTorch框架进行系统测试.实验结果表明,所提方法在准确率、精确率及召回率方面均表现优异,能够较好地捕捉大多数样本.

外文标题：Implementation Method of Speech Recognition System Based on Deep Learning

外文摘要：This article studies the implementation method of speech recognition system based on deep learning.Firstly,explore the overall framework of the speech recognition system.Secondly,delve into the extraction of Mel-Frequency Cepstral Coefficient(MFCC)and the basic principles of Deep Convolutional Neural Network(DCNN).Finally,conduct system testing based on Python and PyTorch frameworks.The experimental results show that the proposed method performs well in accuracy,precision,and recall,and can capture most samples well.

外文关键词：

Deep Convolutional Neural Network(DCNN)speech recognitionPython

作者：

窦亚珍

展开 >

作者单位：

河南农业职业学院,河南郑州 451450

关键词：

深度卷积神经网络(DCNN) 语音识别 Python

出版年：

2024

DOI：