Imagine a computer that doesn't just read text like a robot, but speaks with real human feeling, like excitement or sadness. It's like giving a digital voice a set of emotional acting skills so it can tell stories naturally. This allows a video made in one language to be translated into another while keeping the same emotional punch without needing a recording studio.
Traditional voice dubbing is too expensive and slow for the modern demand for global content. Existing AI text-to-speech tools sound robotic and lack the emotional depth needed to engage audiences.
A generative AI platform called Revoiceit that synthesizes phoneme-level controllable expressive speech using a proprietary HEMM dataset.
If you are a studio dealing with the high cost and time of traditional voice acting for global releases — this project developed a deep learning engine that creates emotionally resonant virtual voices. This allows you to dub content into 16 languages without expensive studio sessions.
If you are an education company dealing with robotic AI voices that bore students — this project developed a platform capable of 32 emotions. This ensures your educational content remains engaging and human-like across different target languages.
If you are an agency dealing with the difficulty of adapting emotional ad campaigns for multiple markets — this project developed a Voice as a Service model. You can now generate expressive virtual voices that maintain the required prosody for 16 different languages.
Based on available project data, specific pricing is not listed, but the project aims to provide a cost-efficient 'Voice as a Service' model to replace expensive studio recording.
Yes, the platform is designed to support more than 32 emotions across 16 different languages.
The technology is based on proprietary deep learning algorithms and a unique proprietary high quality multiemotion multilingual (HEMM) dataset developed by Voiseed SRL.
The project developed user-friendly project management interfaces and released the Revoiceit platform for AI dubbing and production.
The project ran from 2022-01-01 to 2024-06-30, and the enhanced version of the platform (Revoiceit) has already been released.
The project is led by a single Italian SME, Voiseed SRL, representing a 100% industry ratio. This lean structure suggests a direct path from R&D to commercialization without the delays often found in large academic consortia.
Contact Voiseed SRL in Italy
Contact us to explore licensing opportunities for the Revoiceit engine.
