PulseAugur
EN
LIVE 07:03:50

New AI Model Generates Dexterous Robot Grasps with Force Prediction

Researchers have developed EquiDexFlow, a novel SE(3)-equivariant flow-matching model designed to generate dexterous grasps for robotic hands. Unlike previous methods that treat contact forces as a secondary verification step, EquiDexFlow integrates the prediction of wrist pose, joint angles, fingertip contacts, surface normals, and contact forces directly into the model. This approach ensures that predicted grasps are not only kinematically plausible but also physically stable, adhering to contact force conditions and friction constraints by construction. The model has demonstrated strong performance in simulations and on physical hardware, successfully enabling open-loop pick-and-hold trials. AI

RANK_REASON The cluster describes a new academic paper detailing a novel AI model for robotics. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CV →

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →

COVERAGE [1]

  1. arXiv cs.CV TIER_1 English(EN) · Clinton Enwerem, John S. Baras, Calin Belta ·

    EquiDexFlow: Contact-Grounded SE(3)-Equivariant Dexterous Grasp Generative Flows

    arXiv:2606.12728v1 Announce Type: cross Abstract: Most learned dexterous grasp generators relegate contact forces to a downstream verification step, so a kinematically-plausible pose can still violate the conditions for a stable physical grasp. We address this with EquiDexFlow, a…