Nvidia is making its AI-powered Audio2Face tool freely available by open-sourcing it.
This technology creates realistic facial animations for 3D avatars using audio input.
Nvidia is open-sourcing Audio2Face, its AI-driven tool that creates lifelike facial animations for 3D avatars using audio input. This move allows developers to leverage the tool and its framework to build realistic 3D characters for games and applications.
Audio2Face analyzes a voice’s acoustic properties to generate animation data, which it then applies to a 3D avatar’s facial expressions and lip movements. Nvidia highlights that the tool supports both pre-scripted content and real-time livestreams.
Developers, including Farm51 for Chernobylite 2: Exclusion Zone and the team behind Alien: Rogue Incursion Evolved Edition, have already integrated Audio2Face into their games.
Alongside open-sourcing the Audio2Face models and SDKs, Nvidia is releasing the tool’s training framework, enabling users to customize models for specific applications.
Comments
Post a Comment