Demonstration

• Acoustic-to-articulatory inversion using an HMM-based speech production model
• Speaking rhythm conversion by non-negative temporal decomposition