1 paper · 1 filter
Trung Dang, David Aponte, Dung Tran +2
Existing zero-shot text-to-speech (TTS) systems are typically designed to process complete sentences and are constrained by the maximum duration for which they have been trained. H…