Papers by Xincan Feng
Llama-VITS: Enhancing TTS Synthesis with Semantic Awareness (2024.lrec-main)
Copied to clipboard
| Challenge: | Recent advances in Natural Language Processing (NLP) have seen Large-scale Language Models excel at producing high-quality text for various purposes. |
| Approach: | They propose a language model that enriches semantic content of text using Llama2 . their method enhances emotive expressiveness on a dataset . |
| Outcome: | The proposed model matches the naturalness of the original VITS and incorporates BERT (BERT-VITS) on the LJSpeech dataset, highlighting its potential to generate emotive speech. |