Sixth International Conference on Spoken Language Processing
(ICSLP 2000)

Beijing, China
October 16-20, 2000

New Tone Recognition Methods for Chinese Continuous Speech

Jian Liu (1), Tiecheng Yu (2)

(1) Intel China Research Center
(2) Speech Processing Lab., Institute of Acoustics, Chinese Academy of Sciences, Beijing, China

in this paper, we present two tone recognition methods for Chinese continuous speech and the way of integrating methods with the HMM-based speech recognition system. The first method we call it Gaussian Mixture Tone Models (GMTM) method, the second is tone Clustering-Mapping (CM) method. Both of them are statistic methods. At the end of the paper, we will also provide the comparison result of these two methods with the tone classification accuracy tested on the Chinese continuous speech database of national "863 project.


Full Paper

Bibliographic reference.  Liu, Jian / Yu, Tiecheng (2000): "New tone recognition methods for Chinese continuous speech", In ICSLP-2000, vol.1, 377-380.