Estimation of 3D pose from 2D using neural networks
I'd love to see it in action

Open the Pandora's box of motion capture

Using the latest methods and research to advance the possibilities of creating professional level 3D animations from 2D sources.

Capturing movement is hard. It's even harder with limited amount of data like a 2D video. You're missing data on depth or a clear identification of the object you are tracking.

One would think that the task seems impossible, yet advances in AI research have enabled us to create solutions that are capable to provide smooth 3D animations.

It's still not ready for universal application, but we see great advantages when using it for prototyping or for smaller shoots.

Where can our technology help you

Faster Prototyping

Create a prototype animation anywhere. Developers can just use a camera

Shoot preparation

Have and idea how the scene can look like before you start calibrating your optical system.

Small fixes done faster

Avoid the need to book the studio just for a few minutes that need to be fixed.
Want to see
Get in touch with our engineers
Book a meeting