Reproduce “LACE: Latent Visual Representation for Cross-Embodiment Learning” 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
Yoo Sung Jang, Kanchana Ranasinghe, Cristina Mata, Yichi Zhang, Jorge Mendez-Mendez, Michael S. Ryoo
LACE fine-tunes a pretrained visual encoder to align human and robot hand features using sparse keypoint correspondences and a loss that matches relationships among image patches. In real-world “Pick up dino” tests with zero robot task demonstrations, it achieves 60% in-domain and 65% in OOD-env, versus DINO’s 5% and 0%. Results are limited to this grasping setup.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.