利索-全球专利检索

  1. 登录
  2. 注册

搜索结果

2 条记录 (耗时 0.006 秒)

    A CLIENT-SERVER ARCHITECTURE FOR AUTOMATIC SPEECH RECOGNITION APPLICATIONS
    1.
    发明公开
    A CLIENT-SERVER ARCHITECTURE FOR AUTOMATIC SPEECH RECOGNITION APPLICATIONS 审中-公开
    客户 - 服务器架构自动语音识别应用

    公开(公告)号：EP2834812A1

    公开(公告)日：2015-02-11

    申请号：EP13773138.6

    申请日：2013-03-31

    申请人： Dixilang Ltd.

    发明人： SHAGALOV, Victor

    IPC分类号： G10L15/30 ,  G10L15/08 ,  G10L15/20

    CPC分类号： G10L15/083 ,  G10L15/065 ,  G10L15/30 ,  G10L21/0208 ,  G10L25/78

    摘要： A system with aclient-server architecture for Automatic Speech Recognition (ASR) applications, that comprises a client-side including a distributed front end for converting acoustic waves to feature vectors representing the properties of the acoustic waves using DSP; VAD for separating between speech and non-speech acoustic signals and for environmental compensation; adaptor for Web Sockets to pipeline chunked feature vectors to the server side and a server side including A web layer utilizing HTTP protocols and including a Web Server having a Servlet Container for interacting with Java servlets and for managing the lifecycle of servlets using Web Sockets as a transducer that transfers Web Sockets' packets to messages; a Message-Oriented Middleware, connected to the Web Server, for feature vector delivery and decoding and for load balancing between the client side and the server side a recognition server and an adaptation server both connected to the Message-Oriented Middleware, the adaptation server being connected via an adaptation channel, the recognition server interacts with the client side via a recognition channel and a Distributed Frontend (DFE); an intermediate layer for transport, communication and integration; a Speech processing server consisting of a Recognition Server for instantiation of a recognition channel per client and for establishing a communication channel between the web layer and the recognition channel; and a Persistent layer for storing the acoustic models of each client in a hierarchical order, comprising a Language Knowledge Base connected to the recognition server and to the adaptation server, and includes a dictionary, acoustic models, statistical language models and language patterns.