Audio2Face 2022.2 Now Available

This release of A2F 2022.2 running on kit 104.1 brings audio2face to Linux. In addition, a headless mode with a REST api supports a more efficient processing pipeline and export of a large quantity of processed A2F files. Plus, many other improvements to tools found within the a2f application are listed below.

Linux Support
Audio2Face and all its features are now fully supported in a Linux application.

Headless Mode and Rest API
Headless Audio2face supports advanced batch export capabilities by exposing a robust REST api enabling deeper integration into production pipelines.

Multi Blendshape Solve node support
A new solution and interface to allow multiple Blendshape solves and Batch exporting.

Additional updates and features.

Deformer UI - A new UI to allow some high-level management of the deformers employed in a stage.
Warp 0.6.1 - Added support for Cuda Streams and Cuda Events and warp.dll update to Cuda Toolkit 11.5
Mesh correspondence display update - Improved rendering of Point Tags with color and numerical indicators to easily find the tag you are looking for.
CharTransfer head template - A new Default charTransfer head template that closely matches the full-face default.
BlendShape Generation - Now includes the option of ARKit blendshape generator.
Pipeline - Default fields for A2F full face pipeline connections.

Fix - Resolved Hotkeys conflict with default kit hotkeys fixes.
Fix - Mesh separate UV fixes and better mesh handling.
Add - Added undo support for tools.
Fix - Improved Audio file path handling.
Fix - USD Exporter fix.
Fix - Mesh Stamp UV fix.

Audio2Face Video Tutorials
Audio2Face Documentation


Can I ask the $64,000 question. Why is this release not 2023.0?

Seems like this release broke from the year / release pattern that is easy for everyone to understand.

The release date was pushed a few days to accommodate team’s availability to support users, so we counted in New year vacation.

There will be an upcoming 2023 version soon, check back again!

Thank you. I thought your time machines might have had an exact date problem like time machines often have.

I want to ask about the live streaming audio, the demo provided in the tutorial does not seem to meet the real-time continuous pushing, because there will be a pause between each sending, can you help me find the code call of audioplayer living mode

It’s great that A2F now has API support. Is there any documentation about how to use it?

1 Like

That’s a much awaited feature
Looking forward to documentation or sample app with A2F integration.

We are still updating the docs and have a tutorial video on the way for headless mode. In the meantime with the new 2022.2.0 running you can visit http://localhost:8011/docs for interactive API documentation.


The tutorial is now online and it covers how to get started, the documentation, how to run audio2face without UI and some python examples on how to use this API:
Audio2Face Headless and RestAPI Overview - YouTube