Session TimeSunday, 28 July 20191:30pm - 5:30pmMonday, 29 July 201912:15pm - 1:15pmWednesday, 31 July 201912:15pm - 1:15pm
LocationSouth Hall J
DescriptionPOSTER SESSIONS: MONDAY, 29 JULY and WEDNESDAY, 31 JULY, 12:15-1:15 PM
We propose MulTaNet, a deep neural network to solve the multi-task audio-driven facial animation problem in an end-to-end fashion. Extensive experiments indicate that MulTaNet generates more natural-looking and stable facial animation, while showing better generalization capacity for unseen languages compared to previous approaches.