Reproduce “PartialBiGrasp: Inferring Hidden Local Geometry for Bimanual Grasping from Partial Views” 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
Ayush Kaura, Vignesh Vembar, Md Faizal Karim, Keshab Patra, K Madhava Krishna
PartialBiGrasp infers grasp-relevant local geometry from partial observations and uses it to generate and refine coordinated two-arm grasps without reconstructing the complete object at inference. On the unseen DG16M test split, it reports 55.16% FC and 67.87% simulated GS, versus 22.35% and 46.63% for DAGDiff trained on partial point clouds. Physical tests achieved 39/54 after planner filtering.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.