【摘要】 本文介绍了一个用数字信号处理器和十六位微处理器的实时语音识别实验系统的结构.系统采用流水线工作方式,可实现汉语孤立词实时识别.利用透明方式将识别系统与主计算机相联,简化了系统的用户开发和使用.文中还介绍了支持FFT 算法的高速数字信号处理板的结构以及利用通用微机系统实现本系统的开发过程.更多还原

【Abstract】 This paper introduces the structure of an experimental system for real-timespeech recognition composed of a digital signal processor and a 16-bitmicroprocessor.This system operates in the pipeline mode and can recognize isolatedChinese words in real time.The recognition system is connected with the host comput-er transparently,which makes the user employ and develop the system much moresimply.The structure of the high-speed digital signal process board which supportsthe FFT algorithm and the process to develop the system by using a conventionalcomputer and are also described in this paper.更多还原