Figure 13From: Context-dependent acoustic modeling based on hidden maximum entropy model for statistical parametric speech synthesis Preference scores as a subjective comparison between HMEM with decision tree-based context clustering and HSMM. Back to article page