News

"VibeVoice is a novel framework designed for generating expressive, long-form, multi-speaker conversational audio, such as ...
The future wave of innovation will likely be concerned with personalization, enabling readers to personalize the voice, tempo ...
Hume claims Octave is the first text-to-speech system powered by a large language model (LLM) trained not only on text but on speech and emotion tokens, enabling it to understand words in context ...