Self-Supervised Learning for Precise Pick-and-Place Without Object Model

Berscheid, Lars; Meißner, Pascal; Kröger, Torsten

doi:10.1109/lra.2020.3003865

Cited by 64 publications

(30 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…High success rate of novel object pick-and-place based on training-free methods. Among the previous studies on novel object manipulation, few of them have achieved a very high success rate even combined with deep reinforcement learning [3]. But through our method, we have demonstrated by experiments that the idea of similarity matching can lead to an average success rate over 90% with in-category novel…”

Section: Introductionmentioning

confidence: 90%

Category-Association Based Similarity Matching for Novel Object Pick-and-Place Task

Chen,

Kiyokawa,

Wan

et al. 2022

Preprint

View full text Add to dashboard Cite

Robotic pick-and-place has been researched for a long time to cope with uncertainty of novel objects and changeable environments. Past works mainly focus on learningbased methods to achieve high precision. However, they have difficulty being generalized for the limitation of specified training models. To break through this drawback of learning-based approaches, we introduce a new perspective of similarity matching between novel objects and a known database based on categoryassociation to achieve pick-and-place tasks with high accuracy and stabilization. We calculate the category name similarity using word embedding to quantify the semantic similarity between the categories of known models and the target real-world objects. With a similar model identified by a similarity prediction function, we preplan a series of robust grasps and imitate them to plan new grasps on the real-world target object. We also propose a distance-based method to infer the in-hand posture of objects and adjust small rotations to achieve stable placements under uncertainty. Through a real-world robotic pick-and-place experiment with a dozen of in-category and out-of-category novel objects, our method achieved an average success rate of 90.6% and 75.9% respectively, validating the capacity of generalization to diverse objects.

show abstract

Section: Introductionmentioning

confidence: 90%

Category-Association Based Similarity Matching for Novel Object Pick-and-Place Task

Chen,

Kiyokawa,

Wan

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Instead of learning to pick and place objects by using planar manipulation (e.g. a single demonstrated goal state), Berscheid et al [97] trained a robot to pick and place objects by using self-supervised learning without an object model. They combined robot learning of primitives estimated by FCNs and one-shot imitation learning.…”

Section: A Graspingmentioning

confidence: 99%

Review of Deep Reinforcement Learning-Based Object Grasping: Techniques, Open Challenges, and Recommendations

2020

View full text Add to dashboard Cite

The motivation behind our work is to review and analyze the most relevant studies on deep reinforcement learning-based object manipulation. Various studies are examined through a survey of existing literature and investigation of various aspects, namely, the intended applications, techniques applied, challenges faced by researchers and recommendations for minimizing obstacles. This review refers to all relevant articles on deep reinforcement learning-based object manipulation and solutions. The object grasping issue is a major manipulation challenge. Object grasping requires detection systems, methods and tools to facilitate efficient and fast agent training. Several studies have proposed that object grasping and its subtypes are the main elements in dealing with the environment and agent. Unlike other review articles, this review article provides different observations on deep reinforcement learning-based manipulation. The results of this comprehensive review of deep reinforcement learning in the manipulation field may be valuable for researchers and practitioners because they can expedite the establishment of important guidelines.

show abstract

“…In this paper, the target search problem of swarm robots in unknown complex environments is mainly studied, such as forest fire detection (Yao et al, 2018 ; Marzaeva, 2019 ), toxic gas leak detection (Zhang et al, 2010 ; Moshayedi and Gharpure, 2013 ), search and rescue of missing personnel (Goodrich et al, 2009 ; Kamegawa et al, 2020 ), military target detection (Ha and Cho, 2018 ; Jiong et al, 2019 ) and so on. In order to solve this type of search problem, there are mainly composed of two main categories of design strategies, namely, behavior-based search and learning-based search (Cizek and Faigl, 2019 ; Berscheid et al, 2020 ; Suzuki et al, 2020 ), and this article mainly discusses the former.…”

Section: Introductionmentioning

confidence: 99%

Multi-Target Coordinated Search Algorithm for Swarm Robotics Considering Practical Constraints

Zhou

Chen

et al. 2021

Front. Neurorobot.

View full text Add to dashboard Cite

In order to deal with the multi-target search problems for swarm robots in unknown complex environments, a multi-target coordinated search algorithm for swarm robots considering practical constraints is proposed in this paper. Firstly, according to the target detection situation of swarm robots, an ideal search algorithm framework combining the strategy of roaming search and coordinated search is established. Secondly, based on the framework of the multi-target search algorithm, a simplified virtual force model is combined, which effectively overcomes the real-time obstacle avoidance problem in the target search of swarm robots. Finally, in order to solve the distributed communication problem in the multi-target search of swarm robots, a distributed neighborhood communication mechanism based on a time-varying characteristic swarm with a restricted random line of sight is proposed, and which is combined with the multi-target search framework. For the swarm robot kinematics, obstacle avoidance, and communication constraints of swarm robots, the proposed multi-target search strategy is more stable, efficient, and practical than the previous methods. The effectiveness of this proposed method is verified by numerical simulations.

show abstract

Self-Supervised Learning for Precise Pick-and-Place Without Object Model

Cited by 64 publications

References 17 publications

Category-Association Based Similarity Matching for Novel Object Pick-and-Place Task

Category-Association Based Similarity Matching for Novel Object Pick-and-Place Task

Review of Deep Reinforcement Learning-Based Object Grasping: Techniques, Open Challenges, and Recommendations

Multi-Target Coordinated Search Algorithm for Swarm Robotics Considering Practical Constraints

Contact Info

Product

Resources

About