On granularity of prosodic representations in expressive text-to-speech

Re-assign community