Filters








13,894 Hits in 7.2 sec

Can We Learn Finite State Machine Robot Controllers from Interactive Demonstration? [chapter]

Daniel H. Grollman, Odest Chadwicke Jenkins
2010 Studies in Computational Intelligence  
There is currently a gap between the types of robot control polices that can be learnt from interactive demonstration and those manually programmed by skilled coders.  ...  We examine the limits of a regression-based approach for learning an FSM controller from demonstration of a basic robot soccer goal-scoring task.  ...  Introduction We investigate regression-based interactive Robot Learning from Demonstration (RLfD) for use in instantiating a soccer-style goal-scoring Finite State Machine (FSM) controller on a robot dog  ... 
doi:10.1007/978-3-642-05181-4_17 fatcat:zoohodjrjfasjltavzzwrmf2nm

Automata Guided Reinforcement Learning With Demonstrations [article]

Xiao Li, Yao Ma, Calin Belta
2018 arXiv   pre-print
We propose to address these problems by combining temporal logic (TL) with reinforcement learning from demonstrations.  ...  The policy resulting from our framework has an interpretable and hierarchical structure. We validate the proposed method experimentally on a set of robotic manipulation tasks.  ...  φ , D Q ) RLf D stands for any learning from demonstration algorithm Finite state automaton generated from formula ♦(r ∧ ♦(g ∧ ♦b)).  ... 
arXiv:1809.06305v2 fatcat:oe4j2zt53bbuthzfivgi6k4mzq

Incremental Semantically Grounded Learning from Demonstration

Scott Niekum, Sachin Chitta, Andrew Barto, Bhaskara Marthi, Sarah Osentoski
2013 Robotics: Science and Systems IX  
Much recent work in robot learning from demonstration has focused on automatically segmenting continuous task demonstrations into simpler, reusable primitives.  ...  We introduce a novel method for discovering semantically grounded primitives and incrementally building and improving a finite-state representation of a task in which various contingencies can arise.  ...  For this reason, robot learning from demonstration (LfD) [2] has become a popular way to program robots.  ... 
doi:10.15607/rss.2013.ix.048 dblp:conf/rss/NiekumCBMO13 fatcat:msseulkzbbhtngzkgmsbkxn3li

Recent Advances in Robot Learning from Demonstration

Harish Ravichandar, Athanasios S. Polydoros, Sonia Chernova, Aude Billard
2019 Annual Review of Control Robotics and Autonomous Systems  
In the context of robotics and automation, learning from demonstration (LfD) is the paradigm in which robots acquire new skills by learning to imitate an expert.  ...  This review aims to provide an overview of the collection of machine-learning methods used to enable a robot to learn from and imitate a teacher.  ...  (152) used unstructured demonstrations and interactive corrections to learn finite state machines that are made of several trajectory models. Pastor et al.  ... 
doi:10.1146/annurev-control-100819-063206 fatcat:gz56an5s6zh7da7my4ix7gdj4q

Incremental learning of subtasks from unsegmented demonstration

D H Grollman, O C Jenkins
2010 2010 IEEE/RSJ International Conference on Intelligent Robots and Systems  
We illustrate the applicability of our technique by learning a suitable set of subtasks from the demonstration of a finite-state machine robot soccer goal scorer.  ...  Previous robot learning from demonstration techniques have either learned the individual subtasks in isolation, combined known subtasks, or used knowledge of the overall task structure to perform segmentation  ...  Such policies are known to occur in Finite-State-Machine (FSM) controllers [7] .  ... 
doi:10.1109/iros.2010.5650500 dblp:conf/iros/GrollmanJ10 fatcat:hmbggdnzszcwtg2w6ikblcoeyq

Constraint Estimation and Derivative-Free Recovery for Robot Learning from Demonstrations [article]

Jonathan Lee, Michael Laskey, Roy Fox, Ken Goldberg
2018 arXiv   pre-print
Learning from human demonstrations can facilitate automation but is risky because the execution of the learned policy might lead to collisions and other failures.  ...  We present additional conditions, which linearly bound the difference in state at each time step by the magnitude of control, allowing us to prove that the robot will not violate the constraints using  ...  With learning from demonstrations, a robot observes a supervisor policy and learns a mapping from state to control via regression.  ... 
arXiv:1801.10321v3 fatcat:npghadqgnzbpjlkzb76q3hd5k4

Learning Autonomous Mobility Using Real Demonstration Data [article]

Jiacheng Gu, Zhibin Li
2021 arXiv   pre-print
We formulated an architecture based on a long-short-term-memory (LSTM) neural network, which effectively learn reactive control policies from human demonstrations.  ...  Furthermore, we proposed a method to efficiently handle non-optimal demos and to learn new skills, since collecting enough demonstration can be time-consuming and sometimes very difficult on a real robotic  ...  So we train the mobility controller and the manipulation controller separately. A finite state machine is use to control the working models of the robot (Fig.3) . A.  ... 
arXiv:2108.04792v1 fatcat:kbwdz5yspjgvjpo4oa7jj4xk3a

Towards Rapid Multi-robot Learning from Demonstration at the RoboCup Competition [chapter]

David Freelan, Drew Wicke, Keith Sullivan, Sean Luke
2015 Lecture Notes in Computer Science  
Using our method, HiTAB, we can train teams of collaborative agents via demonstration to perform nontrivial joint behaviors in the form of hierarchical finite-state automata.  ...  Finally, we propose a new RoboCup Technical Challenge in multiagent learning from demonstration.  ...  Training an automaton only involves learning its transition functions. In "training mode" the HFA transitions from state to state only when told to by the demonstrator.  ... 
doi:10.1007/978-3-319-18615-3_30 fatcat:23mgwl2iqfhwzccmpxacrns7wu

Learning grounded finite-state representations from unstructured demonstrations

Scott Niekum, Sarah Osentoski, George Konidaris, Sachin Chitta, Bhaskara Marthi, Andrew G. Barto
2014 The international journal of robotics research  
For this reason, learning from demonstration (LfD) has become a popular alternative to traditional robot programming methods, aiming to provide a natural mechanism for quickly teaching robots.  ...  By simply showing a robot how to perform a task, users can easily demonstrate new tasks as needed, without any special knowledge about the robot.  ...  Then, building on these learned skills, we incrementally learn and improve a finite-state representation of a task that allows for intelligent, adaptive sequencing of skills and recovery from errors.  ... 
doi:10.1177/0278364914554471 fatcat:vjgxlppa45g2xp5opisrn4n7su

Learning Sensory-Motor Associations from Demonstration [article]

Vincent Berenz, Ahmed Bjelic, Lahiru Herath, Jim Mainprice
2020 arXiv   pre-print
We propose a method which generates reactive robot behavior learned from human demonstration.  ...  As the experimental section shows, useful behaviors may be learned from a single demonstration covering a very limited portion of the task space.  ...  Next we will investigate how this approach can be used to learn evaluations end-to-end from the raw kinematic states directly and how it could work alongside another learning system which allows for higher  ... 
arXiv:1903.01352v4 fatcat:htcb3daohrhmzihlul4ierybsa

ManiSkill: Generalizable Manipulation Skill Benchmark with Large-Scale Demonstrations [article]

Tongzhou Mu, Zhan Ling, Fanbo Xiang, Derek Yang, Xuanlin Li, Stone Tao, Zhiao Huang, Zhiwei Jia, Hao Su
2021 arXiv   pre-print
Besides supporting the learning of policies from interactions, we also support learning-from-demonstrations (LfD) methods, by providing a large number of high-quality demonstrations (~36,000 successful  ...  We provide baselines using 3D deep learning and LfD algorithms.  ...  building robots, Jiayuan Gu for providing technical support on SAPIEN, and Rui Chen, Songfang Han, Wei Jiang for testing our system.  ... 
arXiv:2107.14483v5 fatcat:5rgpkfezajdidkxk56r5jpj7ri

Incremental Learning Introspective Movement Primitives from Multimodal Unstructured Demonstrations

Hongmin Wu, Zhihao Xu, Wu Yan, Qianxin Su, Shuai Li, Xuefeng Zhou
2019 IEEE Access  
Learning movement primitive from unstructured demonstrations has become a popular topic in recent years, which provides a natural way to endow human-inspired skills to robots.  ...  INDEX TERMS Introspection, movement primitives, unstructured demonstration, Bayesian nonparametric learning, reverse execution, human interaction, kitting experiment.  ...  Consequently, we hopefully learn the robot complex task representation by formulating each state of constructed finite state machine using a state-specific movement primitive technique. 3) LEARNING MOVEMENT  ... 
doi:10.1109/access.2019.2947529 fatcat:hwypvbiz55clrkfwcg5ojkjsyq

Interactive robot task training through dialog and demonstration

Paul E. Rybski, Kevin Yoon, Jeremy Stolarz, Manuela M Veloso
2007 Proceeding of the ACM/IEEE international conference on Human-robot interaction - HRI '07  
We present a framework for interactive task training of a mobile robot where the robot learns how to do various tasks while observing a human.  ...  In this paper, we describe the task training framework, describe how environmental context and communicative dialog with the human help the robot learn the task, and illustrate the utility of this approach  ...  This research was supported by the National Business Center (NBC) of the Department of the Interior (DOI) under a subcontract from SRI International.  ... 
doi:10.1145/1228716.1228724 dblp:conf/hri/RybskiYSV07 fatcat:4nzmohni3zdrrjoyapx4qkvw7m

Learning Constraints from Demonstrations [article]

Glen Chou, Dmitry Berenson, Necmiye Ozay
2019 arXiv   pre-print
We also provide theoretical analysis on what subset of the constraint can be learnable from safe demonstrations.  ...  We extend the learning from demonstration paradigm by providing a method for learning unknown constraints shared across tasks, using demonstrations of the tasks, their cost functions, and knowledge of  ...  Our work is also relevant to human-robot interaction.  ... 
arXiv:1812.07084v2 fatcat:kq4y4jexbzfrrnv7yqit7yuejm

Learning Cooperative Dynamic Manipulation Skills from Human Demonstration Videos [article]

Francesco Iodice, Yuqiang Wu, Wansoo Kim, Fei Zhao, Elena De Momi, Arash Ajoudani
2022 arXiv   pre-print
This article proposes a method for learning and robotic replication of dynamic collaborative tasks from offline videos.  ...  The objective is to extend the concept of learning from demonstration (LfD) to dynamic scenarios, benefiting from widely available or easily producible offline videos.  ...  planner and a Finite State Machine (FSM).  ... 
arXiv:2204.04003v1 fatcat:2fdlyus6zfgvtftuf3kouoowfy
« Previous Showing results 1 — 15 out of 13,894 results