Technical guides and comparisons written by the team building the motion models. We test what we can test, cite our sources, and say where the evidence stops.
Explainers on how motion systems actually work — the mechanisms, the constraints, and the failure modes.
How motion transfers between skeletons and bodies: what IK preserves, why feet slide, and how character retargeting differs from robot retargeting.
Read the guideWhere humanoid training data comes from, how mocap, video, teleoperation and simulation differ, and where human motion stops being sufficient.
Read the guideEvidence-led evaluations of the tools and datasets teams choose between. Where a competitor is the better fit, we say so.
Uthana, DeepMotion, QuickMagic, Rokoko Vision and Cartwheel compared on workflow, exports, pricing, API access and best-fit use case.
Read the comparisonAMASS, HumanML3D, BABEL, KIT-ML, Motion-X, MOSAIC and Uthana by provenance, representation, annotations and licensing.
Read the comparison