Ninth International Conference on Spoken Language Processing

Pittsburgh, PA, USA
September 17-21, 2006

Frequency Warping Based on Mapping Formant Parameters

Zhi-Wei Shuang (1), Raimo Bakis (2), Slava Shechtman (3), Dan Chazan (3), Yong Qin (1)

(1) IBM China Research Lab, China; (2) IBM T.J. Watson Research Center, USA; (3) IBM Haifa Research Lab, Israel

We propose a novel method of generating a frequency warping function by mapping formant parameters of the source speaker and the target speaker. Alignment and selection process are performed to ensure that the mapping formants can represent speakersí difference well. This approach requires only a very small amount of training data for generating the warping function, which can greatly facilitate its application. It can also achieve high quality of the converted speech while successfully converting a speakerís identity. A practical voice morphing system has been built based on this approach. And experimental results show its effectiveness.

Full Paper

Bibliographic reference.  Shuang, Zhi-Wei / Bakis, Raimo / Shechtman, Slava / Chazan, Dan / Qin, Yong (2006): "Frequency warping based on mapping formant parameters", In INTERSPEECH-2006, paper 1768-Thu1BuP.11.