Importing 3D Model with Idle Animations into Audio2Face

AI & Data Science Digital Humans (closed) Audio2Face (closed)
1

Hello everyone,

I am a user of NVIDIA Audio2Face and am currently exploring the capabilities of this amazing software. I’m wondering if it’s possible to import a custom 3D model that already has ‘idle’ animations (resting animations) into Audio2Face. My goal is to synchronize these animations with the facial animation functionality offered by Audio2Face.

Are there specific guidelines or recommended file formats to ensure successful import and integration? Additionally, how does Audio2Face handle pre-existing animations when applying its own facial animations?

I would be very grateful if someone could share their experience or knowledge on this topic.

Thank you in advance for your help!

1 Like
2

Loading usdSkel animations is as easy as loading the stage inside Audio2Face.

Combining face with current body animation while possible is very tricky. But in the upcoming releases it would be much easier.

Here are the steps you should take at the moment.

Part 1

  • Create ArKit blendShapes for your character that has body animation.
  • Load it in avatarstream.bat which is located inside the same folder as Audio2Face.
  • Select the body mesh and from click the menu itemAvatar -> Add BlendShape Receive Setup
  • Check Activate on /World/avatar_stream/ReceiveLivelink
  • Play the timeline animation inside avaterstream.bat

Part 2

  • Open Audio2Face, load one of the example files that has liveLink setup, e.g. mark_solved_arkit
  • Check Activate for the node /World/audio2face/StreamLivelink
  • Play the Audio Player inside Audio2Face. The full body and face animations should play in the avatarstream.bat
3

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.