基于私有云平台的会议语音转写系统  

Conference automatic speech recognition system based on private cloud platform

作　　者：卞飞 王辉[1,2] 穆永鹏 史东东 林林 BIAN Fei;WANG Hui;MU Yongpeng;SHI Dongdong;LIN Lin(The 3RD Research Institute of China Electronics Technology Group Corporation,Beijing 10015,China;Beijing China Electronics Intelligent Acoustics Technology Co.,Ltd.,Beijing 10015,China)

机构地区：[1]中国电子科技集团公司第三研究所,北京100015 [2]北京中电慧声科技有限公司,北京100015

出　　处：《电声技术》2019年第5期1-4,18,共5页Audio Engineering

摘　　要：本文介绍了私有云平台下的语音转写系统的部署和功能特点,从系统架构、语音转写技术、私有云平台和转写音频采集处理器四个方面进行了阐述,指出了在单会议室多设备的情况下实现角色分离软件设计的核心点,介绍了语音转写私有云平台搭建的基本架构,特别说明了语音转写音频采集处理器设计的技术要点产品特点。This paper introduces the deployment and functional characteristics of Automatic Speech Recognition system under private cloud platform.It expounds the system architecture,Automatic Speech Recognition technology,private cloud platform and transcription audio acquisition processor from four aspects.It points out the core points of role separation software design under the condition of single conference room and multi-devices,and introduces the language.The basic architecture of automatic speech recognition private cloud platform is built,especially the technical points and product characteristics of Automatic Speech Recognition audio acquisition processor design.

关 键 词：私有云 语音转写 角色分离 

分 类 号：TN948.63[电子电信—信号与信息处理]