Skip to main content
Figure 2 | EURASIP Journal on Audio, Speech, and Music Processing

Figure 2

From: Lip-Synching Using Speaker-Specific Articulation, Shape and Appearance Models

Figure 2

Two speakers utter here sounds with different make-ups. Colored beads have been glued on the subjects' face along Langer's lines so as to cue geometric deformations caused by main articulatory movements when speaking. Left: a make-up with several hundreds of beads is used for building the shape model. Right: a subset of crucial fleshpoints is preserved for building videorealistic textures.Speaker CDSpeaker OC

Back to article page