Applications
Modern text-to-speech systems produce speech that is
intelligible, but lacking the natural perceptual quality. Significant
improvements in ´naturalness´ of synthetic speech would come with the
ability to adequately control voice source model
- Speaker identification
- Speaker adaptive speech recognition
- Voice morphing
The objective of voice morphing (From the term "metamorphosis", it is to turn
one thing into another) is to
alter the voice characteristics of a "source speaker" so that it is
perceived as to be belonging to a selected "target speaker".
As such voice morphing lends itself to a wide range of applications:
voice editing, karaoke, toys games, speech research tool etc.
|