To accompany the talk on singing synthesis (Track 4, Tuesday 19th 12pm), Chris will be on hand to take participants through CereProc's singing system, including:
- The "Trump" demo - grab a microphone and croon to your heart's content, and prepare to be amazed (horrified) at our beloved POTUS sing right back at you
- STARS transplant procedure - think your singing is bad? Scientifically confirm this first, before allowing us to switch your "bad" singing features with output from our professional singer model
- Vox2XML service - convert a set of vocals and lyrics into CereVoice-compatible singing markup script
- Listen to cutting-edge parametric singing synthesis via CereVoice, harnessing the latest in neural vocoders within speech technology, granting any of our available voices to sing with unparalleled naturalness and expressiveness
This demo session is intended to encourage participants to think about, explore and try out their own ideas of singing interface, as well as be given the opportunity to assess state-of-the-art neural waveform speech and singing synthesis from CereProc's broad range of voices.
