摘要
自动语音识别(Automated speech Recognition)是一种将人的语音转化为文本的技术,它已被运用到了包括教育在内的众多领域,然而在聋儿言语训练领域的应用中却明显滞后。本文通过对采集到的大量聋儿言语信息进行筛选,基于对聋儿语音特点的深入研究,综合国内外先进的语音识别技术,从技术的角度,阐明了言语训练软件原型从前期准备、分析设计到最终架构完成的详细过程,并且创造性地提出了几点架构思路,尤其针对聋儿自身特点,实现了一些针对性强的功能。
Automated Speech Recognition, applied domain of science include education, is a software technology, converting speech into text. But the deaf and dumb speech-training lags behind other domain. The authors filter out a lot of information of the deaf and dumb children speech, research speech trait and integrate speech technology with domestic and international technology. The authors expound the process of prototype architecture of speech-training software, give several creationary thinking, and especially carry out some function with strong points for the deaf and dumb children.
出处
《现代教育技术》
CSSCI
2007年第3期69-73,共5页
Modern Educational Technology
