论文部分内容阅读
语音识别主要是让机器准确地识别出语音的内容。利用隐马尔科夫链(HMM)与人工神经网络(ANN)各自的优点,通过HMM/ANN混合模型处理语音识别有综合的优势。本文就其中的各主要技术展开论述。
Speech recognition is mainly to allow the machine to accurately identify the contents of the voice. Taking advantage of the respective advantages of Hidden Markov Chain (HMM) and Artificial Neural Network (ANN), it is a comprehensive advantage to deal with speech recognition through HMM / ANN hybrid model. This article discusses each of the major technologies.