GraspAgent 1.0: Adversarial Continual Dexterous Grasp Learning
- 19 November 2024
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Robotics and Automation Letters
- Vol. 10 (1) , 366-373
Abstract
Grasp is at the core of robotic manipulation tasks. Nonetheless, most 6-DOF methods resort to a one-time setup via intensive analytics and targeting a predetermined domain. On the other hand, learning and adapting in real environments is of great promise to robotics yet challenging. In this context, this letter presents a grasp learning agent embodied in a dual-arm robot with a gripper and suction ends. Theoretically, GraspAgent is equipped with two adversarial components with the capacity for continual improvement. The first component aims to learn the synthesis of high dexterity grasps using a novel Grasp Awareness Generative Adversarial Network . Thanks to the designed set of contrastive objectives, GA-GAN yields a high rate of feasible grasps for clutter scenes. The second component is an Adversarial Experience Replay . AER promotes certain attributes during the regular rounds of training. For instance, a quality network is forced to adapt the maturity of the sampler by highlighting regions where the sampler performs well. Further, the quality network of each arm is encouraged to learn different grasp primitives, which reflects on better learning capacity. Finally, Experiments on a Yumi robot reveal a final average success rate of 93% after exploiting 17k+ feedback data. Moreover, We show that grasp sampling with GA-GAN surpasses a set of recent baseline methods.Keywords
Funding Information
- Alliance of International Science Organizations, University of Chinese Academy of Sciences
- National Natural Science Foundation of China (91748131, 62006229, 61771471)
- Strategic Priority Research Program of Chinese Academy of Science (XDB32050106)
- InnoHK Project
This publication has 29 references indexed in Scilit:
- Generating quality grasp rectangle using Pix2Pix GAN for intelligent robot graspingMachine Vision and Applications, 2022
- Data-efficient learning of object-centric grasp preferences2022
- A Billion Ways to Grasp: An Evaluation of Grasp Sampling Schemes on a Dense, Physics-Based Grasp Data Set2022
- Graspness Discovery in Clutters for Fast and Accurate Grasp Detection2021
- Robot Learning of 6 DoF Grasping using Model-based Adaptive Primitives2021
- GraspNet-1Billion: A Large-Scale Benchmark for General Object Grasping2020
- 6-DOF GraspNet: Variational Grasp Generation for Object Manipulation2019
- Adversarial Grasp Objects2019
- 4D Spatio-Temporal ConvNets: Minkowski Convolutional Neural Networks2019
- 3D Semantic Segmentation with Submanifold Sparse Convolutional Networks2018