LLM-generated phrase break annotations, produced with only a few prompt examples, rival human annotations in consistency and can train competitive phrase break prediction models in English, French, and Spanish.
Improving prosody with linguistic and bert derived features in multi-speaker based mandarin chinese neural tts,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Synthetic Data Generation for Phrase Break Prediction with Large Language Model
LLM-generated phrase break annotations, produced with only a few prompt examples, rival human annotations in consistency and can train competitive phrase break prediction models in English, French, and Spanish.