12th Annual Conference of the International Speech Communication Association

Florence, Italy
August 27-31. 2011

HMM-Based Emphatic Speech Synthesis Using Unsupervised Context Labeling

Yu Maeno (1), Takashi Nose (1), Takao Kobayashi (1), Yusuke Ijima (2), Hideharu Nakajima (2), Hideyuki Mizuno (2), Osamu Yoshioka (2)

(1) Tokyo Institute of Technology, Japan
(2) NTT Corporation, Japan

This paper describes an approach to HMM-based expressive speech synthesis which does not require any supervised labeling process for emphasis context. We use appealing-style speech whose sentences were taken from real domains. To reduce the cost for labeling speech data with an emphasis context for the model training, we propose an unsupervised labeling technique of the emphasis context based on the difference between original and generated F0 patterns of training sentences. Although the criterion for the emphasis labeling is quite simple, subjective evaluation results reveal that the unsupervised labeling is comparable to the labeling conducted carefully by a human in terms of speech naturalness and emphasis reproducibility.

Full Paper

Bibliographic reference.  Maeno, Yu / Nose, Takashi / Kobayashi, Takao / Ijima, Yusuke / Nakajima, Hideharu / Mizuno, Hideyuki / Yoshioka, Osamu (2011): "HMM-based emphatic speech synthesis using unsupervised context labeling", In INTERSPEECH-2011, 1849-1852.