Investigation of prosodie FO layers in hierarchical FO modeling for HMM-based speech synthesis

To address the overall-micro modeling issue of current prosody model in HMM-based speech synthesis, a hierarchical FO modeling method has been proposed, in which different kinds of pittch patterns are characterized by different prosodie layers and an minimum generation error (MGE) training framework is used to simultaneous optimize FO models of all layers. This paper investigate the importance of prosodie layers and relationship between prosodie characteristics by this hierarchical FO modeling method. Cluster number of each layer is modified to balance the accuracy and robustness of each layer, and thus other layers would be influenced due to the additive structure. The importance and relationship are reflected by different systems with different cluster number ratios. The experimental results and conclusion are valuable and helpful to design a hierarchical FO modeling system.

[1]  Heiga Zen,et al.  Context-dependent additive log f_0 model for HMM-based speech synthesis , 2009, INTERSPEECH.

[2]  Jerome R. Bellegarda,et al.  Statistical prosodic modeling: from corpus design to parameter estimation , 2001, IEEE Trans. Speech Audio Process..

[3]  K. Tokuda,et al.  Speech parameter generation from HMM using dynamic features , 1995, 1995 International Conference on Acoustics, Speech, and Signal Processing.

[4]  Koichi Shinoda,et al.  MDL-based context-dependent subword modeling for speech recognition , 2000 .

[5]  Keiichi Tokuda,et al.  Simultaneous modeling of spectrum, pitch and duration in HMM-based speech synthesis , 1999, EUROSPEECH.

[6]  Ren-Hua Wang,et al.  Minimum Generation Error Training for HMM-Based Speech Synthesis , 2006, 2006 IEEE International Conference on Acoustics Speech and Signal Processing Proceedings.

[7]  Hiroya Fujisaki,et al.  In search of models in speech communication research , 2009, INTERSPEECH.

[8]  Frank K. Soong,et al.  Generating natural F0 trajectory with additive trees , 2008, INTERSPEECH.

[9]  Li-Rong Dai,et al.  Multi-Layer F0 Modeling for HMM-Based Speech Synthesis , 2008, 2008 6th International Symposium on Chinese Spoken Language Processing.

[10]  Frank K. Soong,et al.  A hierarchical F0 modeling method for HMM-based speech synthesis , 2010, INTERSPEECH.

[11]  Keiichi Tokuda,et al.  Hidden Markov models based on multi-space probability distribution for pitch pattern modeling , 1999, 1999 IEEE International Conference on Acoustics, Speech, and Signal Processing. Proceedings. ICASSP99 (Cat. No.99CH36258).