作者: Quingwei Zhao , Qing Guo , Yonghong Yan , Zhiwei Lin , Baosheng Yuan
DOI:
关键词:
摘要: A method and system for generating searching an optimal likelihood decision tree hidden markov model (HMM) based speech recognition are described. Speech signals received. The received processed to generate a plurality of phoneme clusters. clusters grouped into first cluster node second node. determination is made if in the note be moved on increase phone from being