摘要
通过分析MRCPv2协议在语音识别中的典型应用,发现其在实时连续语音识别应用场景中的不足,通过扩展MRCPv2协议解决识别模式区分以及识别内容实时连续上报问题,为电信客服领域辅助语音识别的应用打下基础。
By analyzing MRCPv2 typical applications in speech recognition, it finds its shortage in real-time continuous speech recognition application scenarios. By extending MRCPv2, it solves the identification of recog- nition mode and real-time continuous recognition notification, lays the foundation for customer service in the field of telecommunications assisted speech recognition applications.
出处
《科技创业月刊》
2016年第3期122-124,共3页
Journal of Entrepreneurship in Science & Technology
