Nvidia is letting anyone use its AI voice animation tech

Nvidia is open-sourcing Audio2Face, its AI-powered instrumentality that generates realistic facial animations for 3D avatars — each based connected audio input. The alteration means developers tin present usage the instrumentality and its underlying model to make realistic 3D characters for their games and apps.

Nvidia’s Audio2Face works by analyzing the “acoustic features” of a voice, allowing it to make animation information that it maps to the 3D avatar’s facial expressions and articulator movement. The institution notes that developers tin usage the instrumentality to make 3D characters for pre-scripted content, arsenic good arsenic for livestreams.

Some developers person already utilized Audio2Face successful their games, including Chernobylite 2: Exclusion Zone creator Farm51, arsenic good arsenic the developers of Alien: Rogue Incursion Evolved Edition.

In summation to unfastened sourcing the Audio2Face models and bundle improvement kits, Nvidia is making the tool’s grooming framework available, allowing users to tweak its models for antithetic usage cases.