Reproduce “Reusing Locomotion Training Recipes Across Humanoids via Body-Aligned Virtual Joints” on OpenResearch

Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.

Universal app for Apple silicon and Intel. Requires macOS 11 or later.

About this paper

Sol Choi, Mingon Kim, Seokin Seo, Jongeun Choi, Donghyeon Kim, Yisoo Lee

The authors introduce an analytic body-aligned virtual-joint interface for humanoid locomotion, translating shared anatomical commands into robot-specific joints while retaining existing controllers. In simulation, PND Adam’s mean absolute foot-progression angle was 0.5°, versus 4.8° for native coordinates and 14.2° for task space. On Unitree G1, yaw rate was six times lower than native coordinates, a proxy for heading drift in six rollouts.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.