一种低成本语音识别解决方案

A Low-Cost Speech Recognition Solution

作为人机交互领域重要的研究对象,语音识别技术已经成为信息社会不可或缺的组成部分。目前基于在线引擎和语音芯片实现的语音技术方案,其适用性和使用成本均限制了技术的应用和推广。通过对离线语音识别引擎的研究,结合特定领域内的应用特点,提出一套适用性强,成本较低的语音识别解决方案,可以在离线的网络环境中,实现非特定人的连续语音识别功能。最后根据本方案设计语音拨号软件,并对语音拨号软件的功能进行科学的测试验证。

As an important research topic in the field of human-computer interaction,speech recognition technology has become an indispensable part of the information society.At present,due to the problem of adaptability and cost,the voice technology scheme based on online engine and voice chip is limited.Based on the research of off-line voice engine,combined with the application characteristics in a specific field,this paper proposes a set of solutions with strong applicability and low cost,which can realize the speech recognition function in the off-line environment.Finally,the voice dialing software is designed according to this scheme,and carries on the scientific test to its function.