TransGrasp: Grasp Pose Estimation of a Category of Objects by Transferring Grasps from Only One Labeled Instance

TL;DR

TransGrasp is a novel method enabling category-level grasp pose estimation by transferring grasps from a single labeled object, reducing the need for extensive annotations or detailed models, and improving robotic grasping capabilities.

Contribution

It introduces a shape correspondence-based grasp transfer approach and a refinement module, allowing effective grasp prediction from minimal labeled data.

Findings

Achieves high-quality grasps with transferred poses

Reduces annotation requirements for grasp training

Demonstrates effectiveness on various object categories

Abstract

Grasp pose estimation is an important issue for robots to interact with the real world. However, most of existing methods require exact 3D object models available beforehand or a large amount of grasp annotations for training. To avoid these problems, we propose TransGrasp, a category-level grasp pose estimation method that predicts grasp poses of a category of objects by labeling only one object instance. Specifically, we perform grasp pose transfer across a category of objects based on their shape correspondences and propose a grasp pose refinement module to further fine-tune grasp pose of grippers so as to ensure successful grasps. Experiments demonstrate the effectiveness of our method on achieving high-quality grasps with the transferred grasp poses. Our code is available at https://github.com/yanjh97/TransGrasp.