论文部分内容阅读
文中介绍了一种基于 IP网的顽健语音识别技术 ,首先分析了影响 IP网语音识别技术的两大因素。接着简要介绍了 ITU- T语音编码标准 G. 72 3. 1,比较了传统 IP网语音识别技术和基于解码信号的始端 -尾端语音识别技术 ,并详细介绍了 IP网语音识别技术。最后 ,实验数据表明基于解码信号的始端 -尾端语音识别技术相对于传统 IP网语音识别技术 ,在单字语音识别和连续语音识别的识别率上具有一定的改善
This paper introduces a stubborn voice recognition technology based on IP network, first of all, analyzes two major factors that affect IP voice recognition technology. Then briefly introduces the ITU-T speech coding standard G. 72 3. 1, compares the traditional IP network voice recognition technology and the decoder-based end-end voice recognition technology, and introduces the IP network voice recognition technology in detail. Finally, experimental data show that the beginning-end speech recognition technology based on the decoded signal has a certain improvement over the recognition rate of single-word speech recognition and continuous speech recognition over the traditional IP network speech recognition technology