Llama-VITS: Enhancing TTS Synthesis with Semantic Awareness

arXiv:2404.06714v1 Announce Type: new
Abstract: Recent advancements in Natural Language Processing (NLP) have seen Large-scale Language Models (LLMs) excel at producing high-quality text for various purposes. Notably, in Text-To-Speech (TTS) systems, the integration of BERT for semantic token generation has underscored the importance of semantic content in producing coherent speech outputs. Despite this, the specific utility of LLMs in enhancing TTS synthesis remains considerably limited. This research introduces an innovative approach, Llama-VITS, which enhances TTS synthesis by enriching the semantic …

abstract arxiv bert cs.cl cs.sd eess.as excel importance integration language language models language processing llama llms natural natural language natural language processing nlp processing quality scale semantic speech synthesis systems text text-to-speech token tts type