Skip to main content
Figure 7 | EURASIP Journal on Audio, Speech, and Music Processing

Figure 7

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

Figure 7

Comparing natural (dark blue) and synthetic trajectories computed by three different systems for the first 6main articulatory parameters (jaw opening, lip spreading, jaw protrusion, lower and upper lip opening, laryngeal movements) for the sentence "The lack of employment ensures that the poor earn less than it costs to survive." The three systems are concatenation of audiovisual diphones (black), HMM-based synthesis (light blue), and the proposed PHMM (red). Vertical dashed lines at the bottom of each caption are acoustic boundaries while gestural boundaries are given by the top plain lines. Note the large delay of the non audible prephonatory movements at the beginning of the utterance. The trajectories of lower and upper lips for the word "ensures" is zoomed and commented in Figure 8.

Back to article page