In this paper, the issue of tone recognition for Mandarin speech is discussed. First, a new reliable pitch detection method, which incorporates a sophisticated pitch contour tracking algorithm into the SIFT method, for feature extraction is discussed. Some tone recognition techniques previously proposed using HMM and NN approaches for both isolated-syllable and continuous-speech cases are then reviewed. Lastly, an isolated word recognition system which integrates an MLP tone recognizer with an HMM base-syllable recognizer is discussed.
|Number of pages||11|
|Journal||Journal of the Chinese Institute of Electrical Engineering, Transactions of the Chinese Institute of Engineers, Series E/Chung KuoTien Chi Kung Chieng Hsueh K'an|
|State||Published - 1 Jan 2000|