LogIn
Faceform Neural Profile produces facial motions by looking at two images from the cameras.
Wrap4D creator Faceform keeps perfecting its motion processing workflows, this time by introducing Faceform Neural Profile, a net that can process hours of facial performance. According to the company, it produces facial motions by looking at two images from the head-mounted cameras (HMC).
Wrap4D's motion processing pipeline provides great accuracy but it's computationally intensive, so you are limited to the duration of sequences you can process and can't use it for processing hours of facial performance.
"So we decided to try the following idea: what if we take an HMC and capture a subset of an actor’s performance, say 10-15 minutes of everything the actor is capable of doing with their face, process it with Wrap4D, and train a neural net to produce the same result by looking at just two images from the cameras. We call it Faceform neural profile."
Image credit: Faceform
Faceform demonstrated that you can load the footage from HTC into the Take Manager and pass it to a NeuralNet node, which contains a path to a Faceform neural profile. The facial expression is then transferred to the mesh in just a few seconds.
"We can see the overlap of our resulting mesh with the original camera sequence. And you can see that the tracking provides pretty great accuracy. So having this personalized neural net profile, basically makes the tracking of performance fully automated. And we can literally process hours of facial performance overnight."
Image credit: Faceform
Faceform also mentioned that you might not even need to animate anything as the net captures small animation nuances.
See Faceform Neural Profile in action here and join our 80 Level Talent platform and our Telegram channel, follow us on Instagram, Twitter, and LinkedIn, where we share breakdowns, the latest news, awesome artworks, and more.