Tone recognition of continuous Cantonese speech based on support vector machines
Publication in refereed journal


引用次數
替代計量分析
.

其它資訊
摘要Tone is an essential component for word formation in all tone languages. It plays a very important role in the transmission of information in speech communication. In this paper, we look at using support vector machines (SVMs) for automatic tone recognition in continuously spoken Cantonese, which is well known for its complex tone system. An adaptive log-scale 5-level F-0 normalization method is proposed to reduce the tone-irrelevant variation of F-0 values. Furthermore, an extended version of the above normalization method that considers intonation is also presented. A tone recognition accuracy of 71.50% has been obtained in a speaker-independent task. This result compares favorably with the results reported earlier for the same task. Considerable improvement has been achieved by adopting this tone recognition scheme in a speaker-independent Cantonese large vocabulary continuous speech recognition (LVCSR) task. (C) 2004 Elsevier B.V. All rights reserved.
著者Peng G, Wang WSY
期刊名稱Speech Communication
出版年份2005
月份1
日期1
卷號45
期次1
出版社ELSEVIER SCIENCE BV
頁次49 - 62
國際標準期刊號0167-6393
電子國際標準期刊號1872-7182
語言英式英語
關鍵詞automatic speech recognition; F-O normalization; support vector machines; tone language; tone recognition
Web of Science 學科類別Acoustics; ACOUSTICS; Computer Science; Computer Science, Interdisciplinary Applications; COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

上次更新時間 2020-03-12 於 00:30