Face mesh changing into Clair/Mark

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

Hi,

I have been trying to import custom mesh to audio2face and I am following a tutorial posted by Nvidia.

But whenever I add a Audio pipeline(streaming), the mesh of the character morphs in a way that it looks like Clair or Mark(depend on what I have chosen as a AI model). Does anyone have any advice?

2

It seems you’re missing the warp deformer. Do you have an RTX GPU? If yes, have you installed the latest GPU driver?

Can you take a look at this tutorial to make sure you’re following all the steps?

3

Hi,
Thanks for replying.
I have Nvidia RTX 4060 in my laptop. Unsure if it is enough as I am new to this.

But I found a solution.
After I set up my character, the resemblance happened when I added Streaming pipeline.
But when I add the Regular pipeline then add Streaming pipeline as an extra, my character still looks like mine and gets streaming audio.

Unsure if that was the right workflow or error of streaming pipeline but happy that it works!

Thanks!