Unit selection synthesis in the Smartweb project

Martin Barbisch, Grzegorz Dogil, Bernd Möbius, Bettina Säuberlich, Antje Schweitzer

This paper describes three aspects of the unit selection synthesis used in the SmartWeb dialog system. The synthesis module has been implemented in the IMS German Festival speech synthesis system. First, we compare a unit selection strategy developed in the course of the project to a strategy developed earlier. Second, we discuss our experiences with F0 smoothing and amplitude modeling, which were both devised to reduce audible discontinuities. However, the results are inconclusive so far. Finally, we sketch a simple mechanism that addresses the problem of language disambiguation for proper names.

Cite as: Barbisch, M., Dogil, G., Möbius, B., Säuberlich, B., Schweitzer, A. (2007) Unit selection synthesis in the Smartweb project. Proc. 6th ISCA Workshop on Speech Synthesis (SSW 6), 304-309

@inproceedings{barbisch07_ssw,
  author={Martin Barbisch and Grzegorz Dogil and Bernd Möbius and Bettina Säuberlich and Antje Schweitzer},
  title={{Unit selection synthesis in the Smartweb project}},
  year=2007,
  booktitle={Proc. 6th ISCA Workshop on Speech Synthesis (SSW 6)},
  pages={304--309}
}