A markerless optical motion capture system using computer vision and AI techniques called 4D aim

No restrictive or uncomfortable suits - results in more authentic movement
‍
Excellent results based on input video of sufficient quality: either open sourced; using broadcast video; or filmed using pre-agreed simple set-up and calibration
‍
Self filming is easy - using light and flexible camera systems with minimal set-up and calibration requirements

Eliminates time consuming and expensive mocap studio services with subsequent major data clean-up
‍
Animations can be ‘re-targeted’ to many characters of different shapes and sizes

We call our solution 4D aim: presenting action scenes with multiple authentic humanistic avatars in full natural motion

Fast, high quality authentic versions of real-life movements.

Addresses historical problems of motion capture systems e.g. sensor drift; interference; restrictive suits and hardware; and challenges with object tracking
‍
Technical benefits:

Self-filming using light and flexible camera system with minimal set-up and calibration
100,000+ motion data points captured from 2D video
Track and capture performers and objects separately
Re-target to characters rigs of different shapes and sizes
Direct export into any graphics or game engine

AI Powered Motion Capture