A naukri.com initiative
VentureBeat
1M
68
Image Credit: VentureBeat
Hume launches new text-to-speech model Octave that generates custom AI voices with adjustable emotions
- Hume AI has launched Octave, a new text-to-speech model that can generate custom AI voices with adjustable emotions for various content forms.
- Octave is powered by a large language model trained on text, speech, and emotion tokens to enhance the quality and authenticity of generated voices.
- Users can adjust emotions, tone, rhythm, and cadence of the AI voices on sentence-level using natural language prompts.
- The model can interpret character traits, adjust vocal inflections accordingly, and contextually deliver emotionally accurate speech.
- It supports English and Spanish languages with plans to expand further, catering to content creators in audiobooks, podcasts, video games, and voiceovers.
- Octave allows granular adjustments like expressing nuanced emotions within sentences, enhancing the overall voice modulation.
- Hume's API provides access to the Octave model, enabling up to 50 requests per minute with various text and description length limits.
- The pricing model for Octave TTS is subscription-based and competitively priced, offering tiers from free to Enterprise plans.
- Octave was preferred over a competitor in a blind comparison study for its audio quality, naturalness, and voice matching accuracy.
- The model is trained on extensive language tokens and datasets, enabling it to infer emotions and maintain consistent character voices in long-form content.
Read Full Article
4 Likes
For uninterrupted reading, download the app