Researches And Applications Of Uyghur Speech Recognition Technology In The Smart Home

Traditional appliances,traditional household equipment,have become outdated.So if there is a way to make the old appliances do not change in the case of the use of smartphones to achieve Uighur speech control control how good ah!In this idea I started the research work in this round the intelligent home system is based on the family,using pattern recognition,control instruction mapping,communication network formation,speech recognition,semantic understanding,third-party intelligent control components.At present in the market,there is no Uighur voice control of intelligent home products.Uighur speech recognition technology has matured in recent years,Xinjiang University,led by a group of researchers in the long-term efforts of Xinjiang university to accumulate the relevant fields of key technologies.In the field of Uighur speech recognition research,the College of Information Science and engineering of Xinjiang University has strong scientific research ability and has accumulated over the years.I was involved in the laboratory speech recognition related projects in the graduate stage,and benefited a lot from it.I use three years of study time to integrate the knowledge of speech recognition into the background of intelligent home hardware equipment,the Uighur speech recognition technology applied to the field of intelligent home.Through the speech technology to recognize the integration of technology and intelligent home technology in the old home appliances Uighur speech control,set up a convenient,intelligent home Control client application,making the system more intelligent and humane.This paper focuses on the speech recognition based on the CMU Sphinx 4.0 Uygur Language and the realization of the voice control electrical equipment through the Intelligent Home Hardware interface(TV,light bulb,hot water kettle,water heater,etc.through the switch control of electrical equipment and through remote control of electrical equipment).The speech recognition process of Uighur phrase consists of four parts(speech signal feature parameter extraction,building acoustic model,constructing language model and decoder part pattern matching).This paper makes a detailed comparison and analysis of the four related technologies,chooses MFCC,hmm and N Gram techniques and their related algorithms,and in addition,the paper elaborates the MFCC feature parameter extraction method,hmm acoustic model and its correlation algorithm as well as the N Gram language model and smoothing processing method.Implementation of the language model and acoustic model integrated into the smartphone application in the Android smartphone the speech recognition module of speech control Intelligent Home System final speech the recognition module realizes the control of the electrical equipment through the API of the smart home hardware connected with the mobile Wi Fi Wireless communication protocol and realizes the application of the Uighur speech control intelligent Home.