基于语音识别的智能WIFI音箱

发布时间：2018-08-11 13:16

【摘要】：语音识别技术就是让机器通过识别和理解过程把语音信号转变为相应的文本或命令的技术。基于语音识别芯片的智能硬件产品越来越多。在软件上,目前比较主流的语音识别软件有:Nuance、Google 的 google assistant 和 Microsoft 的 Cortana,Amazon的Alexa,包括国内的科大讯飞的语音识别算法,这些软件都是面向非特定人、大词汇量的连续语音识别系统。硬件上,最早从Amazon的Echo音箱的推出,到国内叮咚音箱,还有刚刚上市的Goolge homer,包括Microsoft也将推出自己的基于语音识别功能的智能音箱。因为智能音箱和和这些语音识别技术的结合很好的解决了智能产品的入口问题,已经在行业形成成共识,不久的将来智能音箱会成为家庭的控制中心.本文对语音识别技术的基本原理进行了初步研究,并分析了语音识别的实现过程和算法介绍。对WIFI音箱系统的本地语音识别方案和实现方法做了说明,并对云端语音识别方案做了介绍。并介绍了 WIFI音箱系统的特点和不同,并对音箱系统的特殊场景的问题进行分析和解决方案的提出。
[Abstract]:Speech recognition is a technology that enables a machine to transform a speech signal into a corresponding text or command through the process of recognition and understanding. There are more and more intelligent hardware products based on speech recognition chip. At present, the popular speech recognition software is google assistant of google assistant and Alexa of Microsoft, including iFLYTEK's speech recognition algorithm, which is a kind of continuous speech recognition system with large vocabulary. Hardware, from the introduction of Amazon's Echo speakers to the domestic Dingdong speaker, as well as the newly launched Goolge homer, including Microsoft will also introduce their own smart speaker based on speech recognition. Because the combination of smart speakers and these speech recognition technologies has solved the problem of entry of intelligent products, there has been a consensus in the industry that smart speakers will become the control center of home in the near future. In this paper, the basic principle of speech recognition is studied, and the realization process and algorithm of speech recognition are analyzed. The local speech recognition scheme and implementation method of WIFI speaker system are described, and the cloud speech recognition scheme is also introduced. The characteristics and differences of the WIFI speaker system are introduced, and the problems of the special scene of the speaker system are analyzed and the solutions are put forward.
【学位授予单位】：东南大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN912.34;TN912.26

【相似文献】