Return to Issue Details Controllable Audio Generation for Virtual Characters Based on Speaker Identity and Expressive Speech Modeling Download Download PDF