The Application Of Tibetan Speech Recognition In Distance Education

Tibetan areas are affected by the geographical location and other factors,the exchange of information is not convenient,education resources is more lacking compared with other regions.The advantages of modern distance education make up for the lack of educational resource in Tibetan schools,and narrow the gap between Tibetan areas' education resource and other areas'.Speech recognition in the distance education platform is an important module of video processing in the network,and it is also one of the most active research topics in the world.This thesis introduces the related principle of Tibetan speech recognition technology and the necessity of applying in distance education.Furthermore,combined with the pronunciation of Lhasa-Tibetan,we produced a lexicon and explored hidden markov model development kit to establish the acoustic model and language model for Tibetan based on phonemes.Then we adopted the acoustic model,language model,pronunciation dictionary and other related data which produced the best performance to the Ubuntu system,and then we used the Qt Creator to develop the Tibetan speech recognition system to realize the functions of real-time sound collection and real-time speech recognition,and finally the results are displayed in the form of Tibetan characters.Finally,this paper applies the system to the Tibetan primary and middle school language teaching video in distance education.Firstly we divide the video into several video segments,and then extract the audio files from the video.Using the Tibetan continuous speech recognition system to identify the Tibetan language in audio files,we can get teaching videos with subtitles through the integration of video files.In this way,we establish the correspondence between videos and texts,which can not only facilitate the users to find video segments,but also improve the efficiency of learning and it is benefit for watching video and understanding knowledge.In the future,users can also search the key words in the teaching videos so that users can use it more conveniently and quickly.