A naukri.com initiative
VentureBeat
5d
365
Image Credit: VentureBeat
Voice AI that actually converts: New TTS model boosts sales 15% for major brands
- Rime has introduced Arcana text-to-speech (TTS) model to create diverse and natural voices, leading to a 15% sales boost for customers like Domino’s and Wingstop.
- The TTS model generates voices based on text descriptions, offering infinite variability across demographics and languages.
- Rime's model enables users to specify desired voice characteristics such as age, gender, and language.
- The company's innovative approach involves training the model on natural conversations with real people instead of voice actors.
- Rime's Mist v2 TTS model is designed for high-volume, business-critical applications, enhancing customer interactions.
- Rime offers eight flagship speakers with unique traits and the ability to switch between languages, adding nuances like laughter and sarcasm.
- The model infers emotions from context and produces varied, realistic outputs using audio tokens decoded into speech.
- Arcana TTS was trained in three stages, incorporating conversational techniques, idiolect, and multilingual code-switching.
- Rime collected naturalistic conversational data by recording real conversations and annotating voices with detailed metadata.
- The company focuses on personalized voices for different applications, providing tools for A/B testing and analytics to optimize voice performance.
Read Full Article
22 Likes
For uninterrupted reading, download the app