Enabling your apps to convert text into synthesized speech is an excellent feature for inclusion and ease of use. Microsoft now has over 200 voices, 129 of which are neural voices – those which sound more natural and real. These neural voices have been created using deep learning to ensure that prosody (rhythms and intonation) and pronunciation is accurate.
The big February announcement was the release of Custom Neural Voice. You can record your own voice, upload it to Custom Neural Voice for training the model and create your very own neural voice. The branding possibility of this feature is huge, in fact, AT&T has already created an AI Bugs Bunny interactive experience with this technology. Check out a video here: AT&T Launches Looney Tunes 5G Experience – YouTube
Due to the potential to misuse this technology, it is available technically, but if you are interested in using it, you must apply (Azure Cognitive Services Application for Gated Services (microsoft.com)).
You can find out more about text-to-speech with Microsoft here: Text to Speech | Microsoft Azure and the documentation here: Text-to-speech overview – Speech service – Azure Cognitive Services | Microsoft Docs and how to get started with custom neural voice, here: Improve synthesis with Custom Voice – Speech service – Azure Cognitive Services | Microsoft Docs.
