The trainable trajectory formation model TD-HMM parameterized for the LIPS 2008 challenge

Gérard Bailly (1) , Oxana Govokhina (1) , Gaspard Breton (2) , Frédéric Elisei (3) , Christophe Savariaux (3)

1 GIPSA-MPACIF - GIPSA - Machines Parlantes, Agents Communicants & Interaction Face-à-face
2 Orange Labs [Cesson-Sévigné]
3 GIPSA-Services - GIPSA-Services

We describe here the trainable trajectory formation model that will be used for the LIPS'2008 challenge organized at InterSpeech'2008. It predicts articulatory trajectories of a talking face from phonetic input. It basically uses HMM-based synthesis but asynchrony between acoustic and gestural boundaries - taking for example into account non audible anticipatory gestures - is handled by a phasing model that predicts the delays between the acoustic boundaries of allophones to be synthesized and the gestural boundaries of HMM triphones. The HMM triphones and the phasing model are trained simultaneously using an iterative analysissynthesis loop. Convergence is obtained within a few iterations. Using different motion capture data, we demonstrate here that the phasing model improves significantly the prediction error and captures subtle ontextdependent anticipatory phenomena

Domaines: Traitement du signal et de l'image [eess.SP]; Traitement du signal et de l'image [eess.SP]

Gérard Bailly, Oxana Govokhina, Gaspard Breton, Frédéric Elisei, Christophe Savariaux. The trainable trajectory formation model TD-HMM parameterized for the LIPS 2008 challenge. Interspeech 2008 - 9th Annual Conference of the International Speech Communication Association, Sep 2008, Brisbane, Australia. pp.561. ⟨hal-00339043⟩