Audio Morphing

Approach: There are two variants of our work: inter-voice morphing and intra-voice morphing. In the intra-voice morphing scenario, a single person’s voice is recorded uttering a wide range of utterances. The speaker’s phones are then morphed in time to generate new utterances of the speaker. We note that intra-voice morphing addresses the same problem that concatenative speech synthesis algorithms address, except that instead of simply re-ordering and concatenating original speech samples, intravoice morphing methods smoothly morph between the different recorded sounds. This will help to reduce the amount of recorded audio needed for speech synthesis systems. For example, current stateof-the-art concatenative speech synthesis systems require around 40 hours of recorded audio in order to produce natural-sounding speech.