“Custom Neural Voices” is a part of Microsoft’s Cognitive Services suite of technologies. It allows users to develop highly realistic, humanlike voices by using Microsoft's groundbreaking neural text-to-speech technology. A user will simply record audio samples and then upload training data. Once the voice is created, it can be used in applications for Text-to-Speech or Voice Assistant.
In addition, once the voice is created, it can be used to speak fluently in 13 other languages:
- German (Germany)
- English (Australia)
- English (Canada)
- English (Canada)
- Spanish (Spain)
- Spanish (Mexico)
- French (Canada)
- French (France)
- Italian (Italy)
- Japanese (Japan)
- Korean (Korea)
- Portuguese (Brazil)
- Chinese (Mandarin, Simplified)
This will provide developers the ability to give their application a voice and persona that can be applied to 13 different languages/dialects. Global companies such as BBC, Swisscom, AT&T and Duolingo have built realistic voices that resonate with their brands.
To learn more about custom voices, as well as getting started tutorials, visit https://speech.microsoft.com/customvoice.
