Release of the platform
Version 1.0 of the FabLang platform is out now. You have now the ability to create a TTS language by just uploading audio files and their transcription on the platform.
Develop new languages, dialects and voices with AI
The recent innovations in AI and the massive deployment of voice technologies have considerably increased user expectations in terms of new usages, performance and personalization. But apart from the major international languages, the linguistic offer remains limited.
Test our solution
35/200 characters
How it works
Data upload : Please upload audio data and their respective transcription. This will be used by our algorithms for TTS system adaptation. Please be careful about the User Confirmation of Rights
Data validation : The data is analyzed and verified automatically and if needed manually.
Model adaptation : Our model is then trained or adapted on the provided data. This step is very dependent of the provided data, it could last from few hours to few days depending on the quality and size of the uploaded data.
Once training is made, the model is available for test in a type and talk interface. Feedback may be given by users.News
FAQ
How can I register ?
Please use the following sign up page and enter a valid email, name, and entity information.
What data is needed to create a new language ?
Two types of data are primarily needed: audio data and their transcriptions, which are required for model training and adaptation. A minimum of 1 hour of data is required to create a system, but naturally, the more data the system receives, the better its performance in terms of naturalness.
Data property
Please ensure that you have the necessary rights and permissions to use the audio and text files uploaded to our system, and that the materials provided do not infringe on any copyrights, intellectual property, or other legal rights of third parties.
What are the available languages ?
For demonstration purposes, only US English is available. However, when logging in, there are no limitations on the languages that can be developed, as long as you have sufficient data to train/adapt the model.
Commercial exploitation
Any commercial use of the generated TTS audio files is not authorized. If you intend to use the files for commercial purposes, please contact us at frontdesk@acapela-group.com for the appropriate licensing arrangements.
