论文部分内容阅读
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.
This paper introduces a real-time speech recognition experimental system with digital signal processor and sixteen microprocessors.The system adopts pipeline working mode to realize real-time recognition of isolated Chinese words.It uses the transparent way to associate the recognition system with the host computer , Which simplifies the development and usage of the system.The paper also introduces the structure of high-speed digital signal processing board that supports FFT algorithm and the development process of the system by using a common microcomputer system.