Hear your face: Multi-speaker speech synthesis with facial inputs only
Affiliation
MARG
Presenter
이재준
Personal Link
Subject
Speech
Site
B3
Time
Poster Session II - 13:30~15:00
1 more property
Abstract
This research aims to multi-speaker speech synthesis that doesn't need any audible input both for contents or speaker embedding. It takes EMG signal as content information and face image for speaker information.
