Deep Imitation Learning for Bimanual Robotic Manipulation

Xie, Fan; Chowdhury, Alexander; Kaluza, M. Clara De Paolis; Zhao, Linfeng; Wong, Lawson L. S.; Yu, Rose

doi:10.48550/arxiv.2010.05134

Cited by 3 publications

(3 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The field of bimanual manipulation has been long studied as a problem involving both hardware design and control [45,21,59,49]. In recent years, researchers applied learning based approach to bimanual manipulation using imitation learning from demonstrations [62,17,54,60] and reinforcement learning [30,1,8,10,18]. For example, Amadio et al [1] proposed to leverage probabilistic movement primitives from human demonstrations.…”

Section: Related Workmentioning

confidence: 99%

DAIR: Disentangled Attention Intrinsic Regularization for Safe and Efficient Bimanual Manipulation

Zhang,

Jian,

et al. 2021

Preprint

View full text Add to dashboard Cite

We address the problem of solving complex bimanual robot manipulation tasks on multiple objects with sparse rewards. Such complex tasks can be decomposed into sub-tasks that are accomplishable by different robots concurrently or sequentially for better efficiency. While previous reinforcement learning approaches primarily focus on modeling the compositionality of sub-tasks, two fundamental issues are largely ignored particularly when learning cooperative strategies for two robots: (i) domination, i.e., one robot may try to solve a task by itself and leaves the other idle; (ii) conflict, i.e., one robot can easily interrupt another's workspace when executing different sub-tasks simultaneously. To tackle these two issues, we propose a novel technique called disentangled attention, which provides an intrinsic regularization for two robots to focus on separate sub-tasks and objects. We evaluate our method on four bimanual manipulation tasks. Experimental results show that our proposed intrinsic regularization successfully avoids domination and reduces conflicts for the policies, which leads to significantly more effective cooperative strategies than all the baselines. Our project page with videos is at https://mehooz.github.io/bimanual-attention/. * Equal contribution.Preprint. Under review.

show abstract

Section: Related Workmentioning

confidence: 99%

DAIR: Disentangled Attention Intrinsic Regularization for Safe and Efficient Bimanual Manipulation

Zhang,

Jian,

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…Alternatively, a number of robotics solutions assume the existence of a manually defined structure, e.g. through specifying an explicit set of primitive skills (Stulp et al, 2012;Xie et al, 2020), or a manually defined curricula (Davchev et al, 2020;. However, hand-crafting structure can be sub-optimal in practice as it varies across tasks.…”

Section: Related Literaturementioning

confidence: 99%

Wish you were here: Hindsight Goal Selection for long-horizon dexterous manipulation

Davchev¹,

Sushkov²,

Regli³

et al. 2021

Preprint

View full text Add to dashboard Cite

Complex sequential tasks in continuous-control settings often require agents to successfully traverse a set of "narrow passages" in their state space. Solving such tasks with a sparse reward in a sample-efficient manner poses a challenge to modern reinforcement learning (RL) due to the associated long-horizon nature of the problem and the lack of sufficient positive signal during learning. Various tools have been applied to address this challenge. When available, large sets of demonstrations can guide agent exploration. Hindsight relabelling on the other hand does not require additional sources of information. However, existing strategies explore based on task-agnostic goal distributions, which can render the solution of long-horizon tasks impractical. In this work, we extend hindsight relabelling mechanisms to guide exploration along task-specific distributions implied by a small set of successful demonstrations. We evaluate the approach on four complex, single and dual arm, robotics manipulation tasks against strong suitable baselines. The method requires far fewer demonstrations to solve all tasks and achieves a significantly higher overall performance as task complexity increases. Finally, we investigate the robustness of the proposed solution with respect to the quality of input representations and the number of demonstrations.

show abstract

“…However, the system becomes more complex as a result because it requires a variety of behaviors, including recovery, to respond robustly to environmental changes (34). One method for integrating multiple combinations of motor primitives using a recurrent graph neural network has been proposed to solve this problem (35). However, the scalability of the primitives is an issue because the recurrent graph network needs to be retrained every time a primitive is added.…”

Section: Introductionmentioning

confidence: 99%

Efficient multitask learning with an embodied predictive model for door opening and entry with whole-body control

et al. 2022

View full text Add to dashboard Cite

Robots need robust models to effectively perform tasks that humans do on a daily basis. These models often require substantial developmental costs to maintain because they need to be adjusted and adapted over time. Deep reinforcement learning is a powerful approach for acquiring complex real-world models because there is no need for a human to design the model manually. Furthermore, a robot can establish new motions and optimal trajectories that may not have been considered by a human. However, the cost of learning is an issue because it requires a huge amount of trial and error in the real world. Here, we report a method for realizing complicated tasks in the real world with low design and teaching costs based on the principle of prediction error minimization. We devised a module integration method by introducing a mechanism that switches modules based on the prediction error of multiple modules. The robot generates appropriate motions according to the door’s position, color, and pattern with a low teaching cost. We also show that by calculating the prediction error of each module in real time, it is possible to execute a sequence of tasks (opening door outward and passing through) by linking multiple modules and responding to sudden changes in the situation and operating procedures. The experimental results show that the method is effective at enabling a robot to operate autonomously in the real world in response to changes in the environment.

show abstract

Deep Imitation Learning for Bimanual Robotic Manipulation

Cited by 3 publications

References 17 publications

DAIR: Disentangled Attention Intrinsic Regularization for Safe and Efficient Bimanual Manipulation

DAIR: Disentangled Attention Intrinsic Regularization for Safe and Efficient Bimanual Manipulation

Wish you were here: Hindsight Goal Selection for long-horizon dexterous manipulation

Efficient multitask learning with an embodied predictive model for door opening and entry with whole-body control

Contact Info

Product

Resources

About