Implementation Of A Mobile Client Based On HMM Speech Recognition Algorithms And IOS Platform

With the continuous development of computer technology and social progress, a very important sign of modern smart devices is the intelligent voice interaction.Speech recognition and speech synthesis inevitably become an important development direction of the current artificial Intelligence research.Speech recognition is a technology that span multiple disciplines.It is mainly based on digital signal processing, and artificial intelligence, as a precondition,involving biology, psychology, linguistics and so many areas.it belongs to pattern recognition areas.Currently, most of the mainstream speech recognition technology is based on the basic theory of the statistical model identification.Hidden Markov models are applied to a lot of good speech recognition system,because of its good performance and robustness.At the same time,with popularity and performance improvement of smart phones,the development of smartphone voice control is very fast,and people need a mobile application without finger control more.In many mobile phone operating system,Apple’s IOS operating system is particularly prominent.From 2007 birth, every year,more and more users are choosing to use IOS operating system.IOS operating system is characterized by its rich aesthetics, robust stability and robust security.Therefore, it is particularly important to develop a mobile App based IOS version.The paper implemented a IOS platform client based on HMM finally,and introduced the author’s ideas in achieving a high recognition rate of speech recognition algorithm.it also introduced some initial ideas for improvement.First of all,This paper,starting from the reality,detailing the status and prospects of development of speech recognition and IOS operating system.Then,it did a detailed discussion and research on the more popular HMM speech recognition technology,detailed study of the algorithm used and did a progressive introduction to IOS operating system develop techniques.Then,in certain technical support,this paper has Completed the HMM speech recognition function by matlab,established a template libraries for speech recognition and has wrote IOS client App by Xcode.The paper also has build a server for interaction.At last,this paper has completed a mobile client based HMM speech recognition algorithm and IOS platforms by combining both them.And it has achieved the ultimate aim and requirements.The paper also introduced the basic concept of the speech recognition,general methods and characteristics of the client and server interaction,HMM implementation of the principles and algorithms and the advantages and application of IOS system.At the same time,this article elaborated the author’s inadequacies of understanding to complete the system and several improved method.