Functional Grasp Transfer Across a Category of Objects From Only one Labeled Instance
- 23 March 2023
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Robotics and Automation Letters
- Vol. 8 (5) , 2748-2755
Abstract
To assist or replace human beings in completing various tasks, research on the functional grasp synthesis of dexterous hands with high degree-of-freedom (DoF) is necessary and challenging. The dexterous functional grasp requires not only that the grasp is stable but more importantly facilitates the functional manipulation after grasping. Such work still relies on manual annotation when collecting data. To this end, we propose a category-level multi-fingered functional grasp transfer framework, in which we only need to label the hand-object contact relationship on functional parts of one object, and then transfer the contact information through the dense correspondence of functional parts between objects, so as to achieve the functional grasp synthesis for new objects based on the transferred hand-object contact information. We verify this method on three categories of representative objects through simulation experiments and achieve successful functional grasps by labeling only one instance in each category.Keywords
Funding Information
- National Natural Science Foundation of China (U1708263)
This publication has 32 references indexed in Scilit:
- UniGrasp: Learning a Unified Model to Grasp With Multifingered Robotic HandsIEEE Robotics and Automation Letters, 2020
- ContactGrasp: Functional Multi-finger Grasp Synthesis from Contact2019
- Generating Grasp Poses for a High-DOF Gripper Using Neural Networks2019
- Multi-step Pick-and-Place Tasks Using Object-centric Dense Correspondences2019
- Learning better generative models for dexterous, single-view grasping of novel objectsThe International Journal of Robotics Research, 2019
- ContactDB: Analyzing and Predicting Grasp Contact via Thermal Imaging2019
- Intel® RealSense™ SR300 Coded Light Depth CameraIEEE Transactions on Pattern Analysis and Machine Intelligence, 2019
- Learning to Grasp Familiar Objects Based on Experience and Objects’ Shape AffordanceIEEE Transactions on Systems, Man, and Cybernetics: Systems, 2019
- Part-based grasp planning for familiar objects2016
- Generating multi-fingered robotic grasps via deep learning2015