语种识别技术在中短波广播强噪声语音音频识别中的应用  

Application of Language Recognition Technology in Medium and Short Wave Broadcasting

作　　者：刘琪 莫东林 LIU Qi;MO Donglin(Sichuan Institute of Industrial Technology,Deyang 618500,China;Chongqing Chuandong Electric Power Group Co.,Ltd.,Chongqing 409100,China)

机构地区：[1]四川工业科技学院,四川德阳618500 [2]重庆川东电力集团有限责任公司,重庆409100

出　　处：《电声技术》2024年第5期49-51,共3页Audio Engineering

摘　　要：在中短波广播中,语音信号经常受到各种噪声的影响,如电磁干扰和环境噪声,增加了语音识别的难度。基于此,全面探讨语种识别技术在中短波广播强噪声环境中的应用,详细介绍特征提取技术、模型适应策略、语言特征库的构建与模型调整方法,旨在提高中短波广播中语种识别的准确率和效率。In medium and short wave broadcasting,speech signals are often affected by various noises,such as electromagnetic interference and environmental noise,which increases the difficulty of speech recognition.Based on this,a comprehensive exploration was conducted on the application of language recognition technology in strong noise environments of medium and short wave broadcasting.Feature extraction technology,model adaptation strategies,construction of language feature libraries,and model adjustment methods were introduced in detail,aiming to improve the accuracy and efficiency of language recognition in medium and short wave broadcasting.

关 键 词：语种识别技术 中短波广播 强噪声语音音频识别 

分 类 号：TN912.34[电子电信—通信与信息系统]