Nvidia unveils AI model for sound generation and modificationed Fugatto, designed for generating music and audio.

This innovative technology can modify voices and produce unique sounds, targeting music, film, and video game producers.

Currently, Nvidia has no plans for a public release of Fugatto.

The model is capable of creating sound effects and music from text descriptions, including unusual sound transformations, like making a trumpet bark like a dog.

Spoken word

What distinguishes Fugatto is its ability to modify existing audio, transforming a piano melody into a sung line or altering a spoken word’s accent and mood.

Bryan Catanzaro, Nvidia’s vice president of applied deep learning research, emphasized the potential of generative AI to enhance music and creative experiences.

The relationship between technology and Hollywood has been complicated, particularly due to concerns over voice imitation, as highlighted by allegations from actress Scarlett Johansson against OpenAI.