Meshcapade grew out of research at the Max Planck Institute on how to represent human bodies in 3D. Its core is SMPL, a parametric model that describes body shape and pose with a small set of numbers, which makes it efficient to generate and animate realistic humans. SMPL has become widely used in computer vision and graphics research, and Meshcapade commercializes the technology.
Its products turn images and video into 3D human data. Motion capture from ordinary video can extract body movement without suits or special cameras, and the output can be used in animation, games, virtual production, and sports analysis. Tools generate avatars with accurate body shapes for virtual try-on in fashion, health and fitness applications, and digital humans. An API lets developers integrate these capabilities, and plugins connect with popular 3D software.
Meshcapade is paid, with plans for individuals and enterprises and licensing arrangements for commercial use of SMPL-based models. Video-based motion capture is convenient but less precise than marker-based studio systems, and results depend on camera angles and lighting. Competitors and alternatives include Move.ai, Rokoko, DeepMotion, Plask, and Wonder Dynamics for video-based animation. It suits teams that want human models grounded in research.