Microsoft’s American voice synthesis female voice Jenny launched on Azure cloud cognitive service

Lately, Microsoft’s American voice synthesis female voice Jenny launched on Azure cloud cognitive service once again made us feel the excellent ability of Microsoft speech synthesis.

Currently, Microsoft has synthesized a total of 129 deep neural network voices, covering 54 languages ​​in the world. Among them, there are 3 American synthetic speeches with as many as 7 different styles. Earlier, Microsoft has released the energetic American male voice Guy and the emotional American female voice Aria.

The recently released American female voice Jenny is kind and gentle. From recording to model training, Jenny’s production is very sophisticated, and its natural sound is very close to real people. The evaluation results show that its MOS (Mean Opinion Score) score is 0.4 points higher than similar products, which is close to the level of real people.

The American-speaking female Jenny has a pure pronunciation, comparable to native speakers. Additionally,  to the default style (General) applicable to various scenarios, Jenny also has styles such as Newscast, Customer Service, Assistant, and Chat.

It can meet the diverse requirements of customers for different tones and styles of pure American language and is suitable for various application scenarios such as translation, reading, dubbing, and teaching.

Microsoft not only pays attention to the quality and naturalness of neural network speech but also pays attention to the diversity of language and speech style. It is hoped that Microsoft can make continuous breakthroughs and innovations in the field of AI speech synthesis, and provide users with more and more high-quality synthetic voice options.

(Source)