INTERSPEECH 2007
8th Annual Conference of the International Speech Communication Association

Antwerp, Belgium
August 27-31, 2007

An Unsupervised Approach to Automatic Prosodic Annotation

Xinqiang Ni (1), Yining Chen (2), Frank K. Soong (2), Min Chu (2), Ping Zhang (1)

(1) Chinese Academy of Sciences, China
(2) Microsoft Research Asia, China

Accent is probably the most prominent part in prosodic events. Automatic accent labeling is important for both speech synthesis and automatic speech understanding. However, manually labeling data for traditional supervised learning is expensive and time consuming. In this paper, we propose an unsupervised learning algorithm to label accent automatically. First, we assume all content words are accented. We build an initial acoustic model with accented vowels in content words and high confidence unaccented vowels in function words. Then an iterative progress is executed to convergence. Experimental results show that this unsupervised learning algorithm achieves about 90% agreement on accent labeling. Compared with 84.3%, the accuracy of a typical linguistic classifier, a 30% relative error reduction is obtained.

Full Paper

Bibliographic reference.  Ni, Xinqiang / Chen, Yining / Soong, Frank K. / Chu, Min / Zhang, Ping (2007): "An unsupervised approach to automatic prosodic annotation", In INTERSPEECH-2007, 486-489.