Zero Resource Speech Challenge 2019: TTS without T
Dear Colleague,
We have the pleasure to announce the new iteration of the Zero Ressource Speech Challenge, which is submitted to Interspeech 2019. Its aim is to build a speech synthesizer without any text or phonetic labels (hence Text to Speech without T). We takes inspiration from young infants who learn to talk before they learn to read or write. Here, the task is to discover a pseudo-text (a sub-word symbolic representation internal to the machine), from raw speech, without any labels, and to use these discovered units to resynthesize new utterances in a target voice.
The organizers
Dunbar, E., Algayres, R., Benjumea, J., Karadayi, J., Cao, X-N., Bernard, M., Ondel, L., Besacier, L., Sakti, S., & Dupoux, E.
Note: The Challenge is a continuation of the
'sub-word unit discovery' task of previous ZeroSpeech challenges, and is
open to everyone (including participants concentrating solely on
sub-word unit discovery or solely on synthesis, as well as participants
building complete end-to-end systems).
--
****************************************
Emmanuel Dupoux
Directeur d'Etudes à l'Ecole des Hautes Etudes en Sciences Sociales
Research Scientist Facebook AI Research
Equipe Cognitive Machine Learning
EHESS - ENS - CNRS - INRIA
29 rue d'Ulm, Paris 75005, France