Reproduce “RoboMMT: Learning Stable Tracking of Modulated Expressive and Interactive Robot Motion” 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
Till Hielscher, Kai O. Arras
RoboMMT stabilizes robots executing expressive motions supplied by a motion library. It trains a tracking policy across bounded modulation of motion references, then uses a separate policy to adjust gesture’s spatial goal. In simulation, it completes at least 99.3% of 2,048 independently sampled 500-step streams on each of three robots, excluding spatial-goal shifts; this measures survival, not accuracy or human-perceived expressivity.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.