Figure 5From: Context-dependent acoustic modeling based on hidden maximum entropy model for statistical parametric speech synthesis Trajectory of log F0 generated from the HSMM, HMEM as well as the natural log F0. Back to article page