Efficient learning of goal-oriented push-grasping synergy in clutter

Xu, Kechun; Yu, Hongxiang; Lai, Qianen; Wang, Yue; Xiong, Rong

doi:10.48550/arxiv.2103.05405

Cited by 2 publications

(19 citation statements)

References 29 publications

(40 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It is however based on deep Q-learning, which is model-free and which does not predict future states. We show in Section VI that our model-based technique significantly outperforms the one from [37] on the same tasks considered in [37] as well as on more challenging ones.…”

Section: Related Workmentioning

confidence: 95%

“…In contrast to these approaches, ours is fully autonomous. The work presented in [37] is most related to ours, with a similar robotic setup and objects. It is however based on deep Q-learning, which is model-free and which does not predict future states.…”

Section: Related Workmentioning

confidence: 99%

“…Similar to studies closely related to the ORC challenge, e.g., [37]- [39], we employ a set of pre-defined and parameterized pushing/grasping manipulation primitives. The decision-making problem then entails the search for the optimal order and parameters of these primitives.…”

Section: B Manipulation Motion Primitivesmentioning

confidence: 99%

“…The objective in previous works is to grasp all the objects; the goal of ORC is to retrieve the target. One straightforward adaption to this new objective is to only give reward when the grasp center is inside the target object, which is the approach that was followed in [37]. However, we found that by providing reward for successfully grasping any object, we can achieve a higher sample efficiency.…”

Section: A Grasp Networkmentioning

confidence: 99%

“…1. VFT is compared with multiple state-ofthe-art approaches [37]- [39], with necessary modifications for solving ORC, i.e., minimizing the number of actions in retrieving a target. The results convincingly demonstrate VFT to be robust and more efficient than the compared approaches.…”

Section: Visual Foresight Tree Search (Vft)mentioning

confidence: 99%

See 4 more Smart Citations

Visual Foresight Trees for Object Retrieval From Clutter With Nonprehensile Rearrangement

Huang

Han

et al. 2022

IEEE Robot. Autom. Lett.

View full text Add to dashboard Cite

This paper considers the problem of retrieving an object from a set of tightly packed objects by using a combination of robotic pushing and grasping actions. Object retrieval from confined spaces that contain clutter is an important skill for robots in order to effectively operate in households and everyday environments. The proposed solution, Visual Foresight Tree (VFT), cleverly rearranges the clutter surrounding the target object so that it can be grasped easily. Rearrangement with nested nonprehensile actions is challenging as it requires predicting complex object interactions in a combinatorially large configuration space of multiple objects. We first show that a deep neural network can be trained to accurately predict the poses of the packed objects when the robot pushes one of them. The predictive network provides visual foresight and is used in a tree search as a state transition function in the space of scene images. The tree search returns a sequence of consecutive push actions that result in the best arrangement of the clutter for grasping the target object. Experiments in simulation and using a real robot and objects show that the proposed approach outperforms model-free techniques as well as model-based myopic methods both in terms of success rates and the number of executed actions, on several challenging tasks.A video introducing VFT, with robot experiments, is accessible at https://youtu.be/7cL-hmgvyec. Full source code will also be made available upon publication of this manuscript.

show abstract

Section: Related Workmentioning

confidence: 95%

Section: Related Workmentioning

confidence: 99%

Section: B Manipulation Motion Primitivesmentioning

confidence: 99%

Section: A Grasp Networkmentioning

confidence: 99%

Section: Visual Foresight Tree Search (Vft)mentioning

confidence: 99%

See 3 more Smart Citations

Visual Foresight Trees for Object Retrieval From Clutter With Nonprehensile Rearrangement

Huang

Han

et al. 2022

IEEE Robot. Autom. Lett.

View full text Add to dashboard Cite

show abstract

Efficient Stacking and Grasping in Unstructured Environments

Wang,

Liu,

Shi

et al. 2024

J Intell Robot Syst

View full text Add to dashboard Cite

Robotics has been booming in recent years. Especially with the development of artificial intelligence, more and more researchers have devoted themselves to the field of robotics, but there are still many shortcomings in the multi-task operation of robots. Reinforcement learning has achieved good performance in manipulator manipulation, especially in grasping, but grasping is only the first step for the robot to perform actions, and it often ignores the stacking, assembly, placement, and other tasks to be carried out later. Such long-horizon tasks still face the problems of expensive time, dead-end exploration, and process reversal. Hierarchical reinforcement learning has some advantages in solving the above problems, but not all tasks can be learned hierarchically. This paper mainly solves the complex manipulation task of continuous multi-action of the manipulator by improving the method of hierarchical reinforcement learning, aiming to solve the task of long sequences such as stacking and alignment by proposing a framework. Our framework completes simulation experiments on various tasks and improves the success rate from 78.3% to 94.8% when cleaning cluttered toys. In the stacking toy experiment, the training speed is nearly three times faster than the baseline method. And our method can be generalized to other long-horizon tasks. Experiments show that the more complex the task, the greater the advantage of our framework.

show abstract

Efficient learning of goal-oriented push-grasping synergy in clutter

Cited by 2 publications

References 29 publications

Visual Foresight Trees for Object Retrieval From Clutter With Nonprehensile Rearrangement

Visual Foresight Trees for Object Retrieval From Clutter With Nonprehensile Rearrangement

Efficient Stacking and Grasping in Unstructured Environments

Contact Info

Product

Resources

About