wrnch debuts latest AR tech at Augmented World Expo 2017

Computer vision start up, wrnch, uses deep learning to turn your smartphone into a motion capture device and find people in the pixels.

Motion capture is the process of digitizing human form and movement. Professional motion capture systems have revolutionized movies, video games, VR, sports analytics and clinical medicine. Up until now, motion capture was expensive, limited to the indoors, not movable and required special equipment and suits.

Wrnch’s BodySLAM is their latest human pose estimation engine to run on a smart phone, which can analyze regular RGB video and returns the 3D motion data for all the humans. Thanks to computer vision and deep learning, BodySLAM allows anyone with a smartphone to capture human motion. The resulting applications are endless ranging from VFX quality camera AR to full body VR to personalized sports analytics.

Key features:

- Real-time interactivity
- 21 body parts tracked per person
- Unlimited number of people in crowded conditions

BodySLAM runs on mobile and PC computing platforms and is fully integrated into world class game engines including UE4 and Unity.

Wrnch is a Montreal-based computer vision start-up, founded by serial AI entrepreneur Dr. Paul Kruszewski and funded by Sharktank Billionaire Mark Cuban.

After months of closed beta, wrnch is officially the BodySLAM engine at Augmented World Expo on June 1st, 2017.