Research Of Small Vocabulary, Speaker-independent Chinese Keyword Spotting Algorithm

Keyword spotting is an important area in speech recognition. Its objective is to identify and verify a few specified key-words in continuous speech. Comparing with keyword spotting, continuous speech recognition needs more resources and its process speed is slower,and it's more vulnerable to noise. So in the art of state,continuous speech recognition is not suitable in many applications and keyword spotting is preferred. If we greatly advance this technology, it will be a great help to expand speech recognition applications.Since the research on keyword spotting of the lab has just started, the foundation of large vocabulary keyword spotting system needs to create dictionary which requires a lot of linguistic knowledge and also needs a big speech database,the paper mainly studies small vocabulary speaker-independent Chinese keyword spotting. In view of the characteristic of Chinese speech, a new keyword spotting algorithm is proposed. In the pre-processing part, continuous speech signal will be divided into syllables through wavelet transform and Teager energy operator. In feature extraction, the MFCC parameters are chosen. Continuous Hidden Markov Model (CHMM) is used to build acoustic model. The decoding process in CHMM is based on Viterbi algorithm, and every two adjacent syllables will be searched together in the recognition process. This method has several advantages. It can improve the search efficiency, reduce the search space and the algorithm complexity. Likelihood ratio test is chosen in keyword verification. All of above achieve the small vocabulary speaker-independent Chinese keyword spotting.Speech signals were got in laboratory. The results of training and testing confirmed the feasibility and validity of the method in this dissertation, and it will play an important role in small vocabulary speaker-independent Chinese keyword spotting.