@inproceedings{zhao2005refining,
author = {Zhao, Yong and Wang, Lijuan and Chu, Min and Soong, Frank and Cao, Zhigang},
title = {Refining Phoneme Segmentations Using Speaker-Adaptive Context Dependent Boundary Models},
booktitle = {INTERSPEECH 2005},
year = {2005},
month = {September},
abstract = {Consistent phoneme segmentation is essential in building high quality Text-to-Speech (TTS) voice fonts. In this paper we propose to adapt an existing well-trained Context Dependent Boundary Model (CDBM) for refining segment boundaries to a new speaker with limited, manually segmented data. Three adaptation approaches: MLLR, MAP, and a combination of the two, are studied. The combined one, MLLR+MAP, delivers the best boundary refinement performance. In comparison with other boundary segmentation methods, the adapted CDBM yields better results, especially with a limited amount of adaptation data. Given 400 manually segmented boundary tokens in about 20 sentences as a development set, the segmentation precision can reach 90% of human labeled boundaries within a tolerance of 20 ms.},
publisher = {International Speech Communication Association},
url = {http://approjects.co.za/?big=en-us/research/publication/refining-phoneme-segmentations-using-speaker-adaptive-context-dependent-boundary-models-2/},
edition = {INTERSPEECH 2005},
}