Nvidia features artificial voice that sounds like natural

The voice controlled by AI will bring rhythm and intonation

Nvidia showed its solution to its boring GPS voice: expressive voice synthesis is an element of NVIDIA’s research work in conversational AI, and tries to bring more natural expressiveness to computer-generated speech.

People speak with a complex rhythm, intonation and timbre that is challenging for the AI ​​to produce. To support developers and developers, researchers in the NVIDIA Enterprise division are building models and tools for high-quality, controllable speech synthesis that captures the richness of human speech without audio artifacts.

This solution makes it possible to use these automated voice models in services such as automated banking and store services, or bringing characters from games and books to life. Below is a sample of the voice creation process and the result:

In addition to expressive speech synthesis, Nvidia’s research group also deals with other elements of Artificial Intelligence (AI) conversation, such as natural language processing, automated speech recognition, keyword detection, audio enhancement, and more. . As it should be, the solutions developed by the working group are accelerated by Nvidia hardware, and their open source can be accessed on the NVIDIA NeMo toolkit, with resources available on the Nvidia NGC HUB.

The expressive voice is also part of a set of technologies under development and exhibited in the series “I am AI”, which in free translation would read “I am the Artificial Intelligence”. Another very impressive demonstration was done at SIGGRAPH, where an avatar is created in real-time, using just a photo as a reference and using the person’s speech and movement to animate the model.

– Continues after advertising –

Interspeech 2021 is an event focused on the latest research and technologies in the field of speech processing. In addition to the expressive voice synthesizer, Nvidia will publicize its participation in the event through this link. You can find out more about Interspeech 2021 on the event’s official website.


Are you thinking of buying a product online? Discover the Adrenaline Save extension for Google Chrome. It is free and offers you price comparisons in the main stores and coupons so you can always buy at the best price. Download now.

Source: Nvidia