Call for Participation: Voice Conversion Challenge 2016
#We apologize for multiple postings.
We are glad to invite you to participate in the first
"Voice Conversion Challenge" that we are organizing soon.
The purpose of this challenge is to compare different systems
and approaches on the task of converting five source speakers'
voices to five different target speakers' voices (females and
males included) from a fixed corpora (both training and
evaluation sets will be provided). In total, 25 conversion
cases will be evaluated in terms of objective audio quality
measures and perceived naturalness and similarity (listening tests).
In this first challenge we focus on 16 kHz speech and
signal-to-signal conversion strategies. No transcription will be
provided and the use of phonetic information is NOT allowed.
However, participants are free of using additional data (for
training purposes) different from the database from which
the speakers corpora will be provided.
The intended schedule is as follows:
- November 23rd: release of training data
(and notification of participation deadline)
- January 4th: release of evaluation data
- January 11th: deadline to submit the converted audio
- February 25th-29th: notification of results
There is no fee for registration, COLIPS
(http://www.colips.org/ <http://www.colips.org/>) has generously
provided financial support to this challenge. We invite however
the participants to notify your intention to participate by
*** November 23rd ***. The challenge guidelines that
the participants should strictly follow will be provided in detail
when releasing the training data.
Please contact us here for questions and notification of
Lingui Chen (University of Science and Technology of China)
Daisuke Saito (Tokyo University)
Tomoki Toda (Nagoya University)
Fernando Villavicencio (National Institute of Informatics)
Zhizheng Wu (University of Edinburgh)
Junichi Yamagishi (National Institute of Informatics)
Looking forward to hear from you.
Voice Conversion Challenge 2016 Organizers