PRESENT: Zero-Shot Text-to-Prosody Control
Lam, Perry, Zhang, Huayun, Chen, Nancy F., Sisman, Berrak, Herremans, Dorien
–arXiv.org Artificial Intelligence
Current strategies for achieving fine-grained prosody control in speech synthesis entail extracting additional style embeddings or adopting more complex architectures. To enable zero-shot application of pretrained text-to-speech (TTS) models, we present PRESENT (PRosody Editing without Style Embeddings or New Training), which exploits explicit prosody prediction in FastSpeech2-based models by modifying the inference process directly. We apply our text-to-prosody framework to zero-shot language transfer using a JETS model exclusively trained on English LJSpeech data. We obtain character error rates (CER) of 12.8%, 18.7% and 5.9% for German, Hungarian and Spanish respectively, beating the previous state-of-the-art CER by over 2x for all three languages. Furthermore, we allow subphoneme-level control, a first in this field. To evaluate its effectiveness, we show that PRESENT can improve the prosody of questions, and use it to generate Mandarin, a tonal language where vowel pitch varies at subphoneme level. We attain 25.3% hanzi CER and 13.0% pinyin CER with the JETS model. All our code and audio samples are available online.
arXiv.org Artificial Intelligence
Aug-13-2024
- Country:
- Europe > Russia (0.04)
- North America > United States
- Texas > Dallas County
- Richardson (0.04)
- Hawaii > Honolulu County
- Honolulu (0.04)
- Texas > Dallas County
- Asia
- Genre:
- Research Report (0.50)
- Technology:
- Information Technology > Artificial Intelligence
- Machine Learning (1.00)
- Natural Language > Large Language Model (0.93)
- Speech > Speech Synthesis (0.59)
- Information Technology > Artificial Intelligence