Nvidia launches AI technology that transforms sound

This AI technology can create unique effects such as turning the sound of a trumpet into a dog barking, adding singing to a melody... based on text descriptions.

Innovative AI technology promises to bring new possibilities to music. (Source: Nvidia)

Date 25/11, Nvidia – the leading company world in the field of chips and software for artificial intelligence (AI) systems, has introduced a breakthrough technology called “Fugatto”. This is an AI model that is capable of not only creating music but also creatively alter and transform sounds, opening up groundbreaking possibilities for creators in the music, film and video game industries.

According to Nvidia, Fugatto can create unique sound effects based on a text description. For example, the technology can turn the sound of a trumpet into a dog barking or add emotional singing to a piano melody, change the voice in a recording from a British accent to an American accent, or even change the mood expressed in a voice, from happy to sad to excited.

“Looking back over the past 50 years, music has changed a lot thanks to computers and synthesizers,” said Bryan Catanzaro, vice president of applied deep learning research at Nvidia. “I believe that creative AI will bring new possibilities to music, video games, and even ordinary people who want to create unique works of art.”

Nvidia does acknowledge, however, that there are risks with the technology. “There is always a risk of misuse with creative technology, such as creating false information or copyright infringement,” Catanzaro warned. “People could use this technology to fake their voices or create pirated characters.”

Fugatto is not the only technology in the field of sound generation and video with AI. Large companies like Meta and startups like Runway are also developing similar models. However, like OpenAI and Meta, Nvidia has not announced a specific time for releasing Fugatto to the market.