[1]张永强,朱海鑫,周万珍,等.单通道频域语音分离技术研究进展[J].计算机技术与发展,2022,32(S2):8-15.[doi:10. 3969 / j. issn. 1673-629X. 2022. S2. 002]
　ZHANG Yong-qiang,ZHU Hai-xin,ZHOU Wan-zhen,et al.Research Progress of Single Channel Frequency Domain Speech Separation Technology[J].,2022,32(S2):8-15.[doi:10. 3969 / j. issn. 1673-629X. 2022. S2. 002]
点击复制

单通道频域语音分离技术研究进展()
分享到：

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:
32
期数:
2022年S2期
页码:
8-15
栏目:
综述
出版日期:
2022-12-11

文章信息/Info

Title:
Research Progress of Single Channel Frequency Domain Speech Separation Technology
文章编号:
1673-629X(2022)S2-0008-08
作者:
张永强1; 2 ; 朱海鑫1 ; 周万珍1 ; 满梦华2
1. 河北科技大学 信息科学与工程学院,河北 石家庄 050018;
2. 陆军工程大学石家庄校区电磁环境效应重点实验室,河北 石家庄 050003
Author(s):
ZHANG Yong-qiang1; 2 ; ZHU Hai-xin1 ; ZHOU Wan-zhen1 ; MAN Meng-hua2
1. School of Information Science and Engineering,Hebei University of Science and Technology,Shijiazhuang 050018,China;
2. National Key Laboratory on Electromagnetic Environment Effects,Shijiazhuang Campus,Army Engineering University,Shijiazhuang 050003,China
关键词:
语音分离技术; 语音信号; 语音信号质量; 深度学习; 语音信号处理
Keywords:
speech separation technology; speech signal; speech signal quality; deep learning; speech signal processing
分类号:
TP181;TN912. 3
DOI:
10. 3969 / j. issn. 1673-629X. 2022. S2. 002
摘要:
为了使人们在语音交互过程中不被干扰,获得准确的信息,语音分离技术应运而生,该项技术是一个热点和难点问题。 语音分离技术能够从嘈杂的环境中将感兴趣的语音提取出来,以此来提高语音信号的信号和感知两个方面的效果,是语音信号处理的重要组成部分。 根据调研国内外学者在语音分离技术方面的研究成果,将语音分离技术归纳为两大类,包括基于传统的语音分离技术和基于深度学习的语音分离技术;简述了各类方法的实现原理、实现流程和所针对的问题,并对不同类别方法之间的优劣进行了对比分析。 除此之外,将与语音分离技术相关的常用数据集分中英文两类进行了归纳总结。 最后,根据语音分离技术目前面临的挑战,对未来的研究发展方向进行展望。
Abstract:
In order to make people undisturbed and obtain accurate information in the process of voice interaction, speech separationtechnology arises at the historic moment,which is a hot and difficult problem. Speech separation technology can extract interested speechfrom noisy environment,so as to improve the effect of signal and perception of speech signal. It is an important part of speech signal processing. According to the research results of domestic and foreign scholars on speech separation technology,speech separation technologyis classified into two categories, including speech separation technology based on traditional speech separation technology and speechseparation technology based on deep learning. We briefly describe the implementation principle,implementation process and problems ofeach method,and compare and analyze the advantages and disadvantages of different categories of methods. In addition,the commonlyused data sets in speech separation technology are divided into two categories:Chinese and English. Finally,according to the current challenges of speech separation technology,the future research and development direction is prospected.

相似文献/References:

[1]韩志艳,王健. 多模式情感识别特征参数融合算法研究[J].计算机技术与发展,2016,26(05):27.
　HAN Zhi-yan,WANG Jian. Research on Feature Fusion Algorithm for Multimodal Emotion Recognition[J].,2016,26(S2):27.
[2]韩志艳,王健. 基于共振峰曲线的语音信号动态特征提取方法[J].计算机技术与发展,2017,27(06):72.
　HAN Zhi-yan,WANG Jian. Dynamic Feature Extraction for Speech Signal Based on Formant Curve[J].,2017,27(S2):72.
[3]张其进,张玉梅.基于混沌特性的语音信号分类[J].计算机技术与发展,2019,29(01):66.[doi:10. 3969 / j. issn. 1673-629X. 2019. 01. 014]
　ZHANG Qi-jin,ZHANG Yu-mei.Classification of Speech Signal Based on Chaotic Characteristics[J].,2019,29(S2):66.[doi:10. 3969 / j. issn. 1673-629X. 2019. 01. 014]

更新日期/Last Update: 2022-10-10