| On online collaborative learning platform, speech recognition of videos is crucialto facilitate users in faster searching and choosing multimedia files in a network andcompleted the information retrieval operation in teaching videos. The teaching videospeech recognition system in the paper, which is an application case of hidden Markovmodel in extracting texts from audio or video teaching files, has important applicationvalue.This paper reviewed and studied the development and status of speech recognition,and analyzed the relevant principles and technology of speech recognition. Afteranalyzing speech signal preprocessing procedure and comparing the common methodsof extracting feature parameters, the paper built a acoustic model of audio signalrecognition in educational videos and designed the recognition algorithm depended onthe HMM mathematical model. Besides, a keyword table in a special area was made andthe keyword speech recognition was developed by frequently-used kewordrecognization methods of speech signal. Finally, the requirements analysis of appliedsoftware for speech recognition was made, and the performance testing related functionswas completed. |
PDF Full Text Request