It is one of the main actors in the field of artificial intelligence, both with its hardware and models. NVIDIA, “Swiss army knife for audio” a new experimental generative artificial intelligence model, which he describes as Fugatto (Foundational Generative Audio Transformer Opus 1) introduced.

Fugatto is a name in our language. fugue The name of the composition model consisting of polyphonic and repetitive melodies that we call To Fugato It’s very similar. This is not without reason, the artificial intelligence model is also “polyphonic” and different accents and different languages It was developed by researchers from all over the world to detect

Will understand sounds the same way humans do

Making statements about the subject Rafael Valle, NVIDIA applied audio research manager, “We wanted to create a model that understands sounds the way humans understand and produce them.NVIDIA also included Fugatto’s possible use cases in the announcement. These scenarios include prototyping a song idea and trying out different styles, sounds and instruments.

Users can also use this model with with different voice tones and accents, game developers can get help from the model to produce variations on the same sound. Furthermore, with minor modifications, this model can be used in areas where it has not been trained before.

Fugatto public access There was no mention of whether it would be opened or not. Meta and Google had previously developed similar artificial intelligence models.