Reproduce “Precise SE(3) End-Effector Tracking in Whole-Body Humanoid Control” 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
Joohwan Seo, Xiaofeng Guo, Jinkun Cao, Roberto Horowitz, Rocky Duan, Guanya Shi, Koushil Sreenath
ResGAC combines geometric end-effector feedback with a learned policy that adds bounded arm corrections and controls other joints. A ground-attached heading frame reduces reference-induced hand motion from pelvis roll, pitch, and heave. Across four standing trajectories, it had the lowest reported position and orientation errors; pick-and-place position error was 6.6 ± 2.7 mm. The tracking benchmarks were standing tests.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.