Phoneme Based Speaker-Independent English Command Recognition

In this paper we propose a new algorithm of phoneme based speaker-independent English command recognition and develop a speaker-independent English command recognition system. It accelerates the whole system development by using HTK (hide Markov toolkits) and Visual C++ based on the character’istics of speaker-independent speech recognition. In recognition phase we combine the confidence measures with incomplete matching, which considerably improve the quality of recognition. The recognition accuracy is increased by 4.8% over complete matching without back-end processing when the sige of vocabulary is more than 10.