Filters








6,130 Hits in 4.5 sec

Learning an internal representation of the end-effector configuration space [article]

Alban Laflaquière, Alexander V. Terekhov, Bruno Gas, J.Kevin O'Regan
2018 arXiv   pre-print
An internal representation of the end-effector configuration is generated from unstructured proprioceptive and exteroceptive data flow under very limited assumptions.  ...  A mapping from the proprioceptive space to this representational space can then be used to control the robot.  ...  The goal of the robot is to create an internal representation ξ of the endeffector configuration using topological information of the exteroceptive space. space of end-effector configurations C.  ... 
arXiv:1810.01866v1 fatcat:xwklqtgqs5ekjh2vacnz57nrjy

A Self-Organizing Neural Model of Motor Equivalent Reaching and Tool Use by a Multijoint Arm

Daniel Bullock, Stephen Grossberg, Frank H. Guenther
1993 Journal of Cognitive Neuroscience  
The Need for Internal Spatial Representations Several different phenomena fall under the general heading of motor equivalence.  ...  Central to the model is an analysis of how visual, spatial, and motor representations are formed and combined for the control of goal-oriented reaching.  ...  As noted above, both internal and external feedback loops exist for updating the internal representation of end effector position.  ... 
doi:10.1162/jocn.1993.5.4.408 pmid:23964916 fatcat:45giaas6jbf6np3oqsf4jhmuxi

Population based Mean of Multiple Computations networks: A building block for kinematic models

Manuel Baum, Martin Meier, Malte Schilling
2015 2015 International Joint Conference on Neural Networks (IJCNN)  
Such a representation will be introduced and applied for the case of a redundant manipulator.  ...  The local transformations in between the kinematic variables can be sufficiently well learned by small single MLP layers.  ...  This is achieved by shaping the models attractor space to represent valid combinations of joint-configurations and end-effector positions.  ... 
doi:10.1109/ijcnn.2015.7280791 dblp:conf/ijcnn/BaumMS15 fatcat:vuf5waidyvc35bgotb6b2fdqke

Learning local linear Jacobians for flexible and adaptive robot arm control

Patrick O. Stalph, Martin V. Butz
2011 Genetic Programming and Evolvable Machines  
We show that XCSF can learn large forward velocity kinematic mappings autonomously and rather independently of the task space representation provided.  ...  Successful planning and control of robots strongly depends on the quality of kinematic models, which define mappings between configuration space (e.g. joint angles) and task space (e.g.  ...  Open Access This article is distributed under the terms of the Creative Commons Attribution Noncommercial License which permits any noncommercial use, distribution, and reproduction in any medium, provided  ... 
doi:10.1007/s10710-011-9147-0 fatcat:5gruwdjiy5bmlfkr2wcdpvfpyy

Representation and generalization of bi-manual skills from kinesthetic teaching

Rene Felix Reinhart, Andre Lemme, Jochen Jakob Steil
2012 2012 12th IEEE-RAS International Conference on Humanoid Robots (Humanoids 2012)  
Skills are learned and embedded over several representational levels comprising a compact movement representation by means of movement primitives, a task space description of the bi-manual tool constraint  ...  By means of comparative evaluation of different architectural configurations, a specific modulation of skill execution with respect to the embedding in the robot's workspace is identified to achieve optimal  ...  An automated data analysis is conducted to determine the task space coordinates g of the guiding hand, i.e. the end effector trajectory with the highest spatial variance.  ... 
doi:10.1109/humanoids.2012.6651575 dblp:conf/humanoids/ReinhartLS12 fatcat:7l43q4mga5hmlc6wm7rv74ttfq

Transfer of knowledge for a climbing Virtual Human: A reinforcement learning approach

B. Libeau, A. Micaelli, O. Sigaud
2009 2009 IEEE International Conference on Robotics and Automation  
In this paper we present a transfer framework adapted to the case of a climbing Virtual Human (VH). We show that our VH learns faster to climb a wall after having learnt on a different previous wall.  ...  In the reinforcement learning literature, transfer is the capability to reuse on a new problem what has been learnt from previous experiences on similar problems.  ...  The authors would like to thank Cyrille Collette (CEA/LIST) for his help in the use of the VH and ENSTA ParisTech for granting access to their machines.  ... 
doi:10.1109/robot.2009.5152553 dblp:conf/icra/LibeauMS09 fatcat:d6wevn3sa5fy7g3fqsccm5bd2u

Building a Sensorimotor Representation of a Naive Agent's Tactile Space

Valentin Marcel, Sylvain Argentieri, Bruno Gas
2017 IEEE Transactions on Cognitive and Developmental Systems  
The presented work is based in this idea, and a method for the building of an internal representation of its sensorimotor interaction is proposed.  ...  Moreover, plausibility of the internal sensorimotor representation is highlighted by showing that it is possible to consider motion planning directly from it.  ...  an internal representation of the interaction between an agent with its own body.  ... 
doi:10.1109/tcds.2016.2617922 fatcat:kiexyjclofffjpouva2bz3udoy

Learning Global Inverse Statics Solution for a Redundant Soft Robot

Thomas George Thuruthel, Egidio Falotico, Matteo Cianchetti, Federico Renda, Cecilia Laschi
2016 Proceedings of the 13th International Conference on Informatics in Control, Automation and Robotics  
The results indicate that learning based approaches could be an effective method for modelling and control of complex soft robots, especially for high dimensional redundant robots.  ...  A unique inverse statics formulation and data sampling method enables the learning system to circumvent the main roadblocks of the inverting problem.  ...  of Foreign Affairs and International Cooperation DGSP-UST for the support through Joint Laboratory on Biorobotics Engineering project.  ... 
doi:10.5220/0005979403030310 dblp:conf/icinco/ThuruthelFCRL16 fatcat:hyk25du2y5bhlbmo5wrzui6hxa

Training a spiking neural network to control a 4-DoF robotic arm based on Spike Timing-Dependent Plasticity

Alexandros Bouganis, Murray Shanahan
2010 The 2010 International Joint Conference on Neural Networks (IJCNN)  
Its aim is to provide the joint commands that will move the end-effector in a desired spatial direction, given the joint configuration of the arm.  ...  The architecture is a feed-forward network where the input layers encode the intended movement direction of the end-effector in spatial coordinates, as well as the information that is given by proprioception  ...  ACKNOWLEDGMENT This work was supported by an EPSRC Grant under Project Code EP/F033516/1.  ... 
doi:10.1109/ijcnn.2010.5596525 dblp:conf/ijcnn/BouganisS10 fatcat:egbzagura5d6jg6g7ylpq5g3vm

A neural controller for collision-free movement of general robot manipulators

Graf, LaLonde
1988 IEEE International Conference on Neural Networks  
program employing these equations is written to generate arm c o n Q " sequences that move the robot's end-effector from its current position to a target position.  ...  In this paper we show how the mhitectum is able to learn to gsnerate, in constant time, many altanative solution configurations which touch a target point, and we show how leamed collision constraints  ...  During a preliminary training phase, it must learn to associate eye angle vectors with one or more target arm configurations that place the end-effector within an allowable distance from the target point  ... 
doi:10.1109/icnn.1988.23831 dblp:conf/icnn/GrafL88 fatcat:t5nb2htw4bd4fecthwubg3dm4y

Topology-based representations for motion planning and generalization in dynamic environments with interactions

Vladimir Ivan, Dmitry Zarubin, Marc Toussaint, Taku Komura, Sethu Vijayakumar
2013 The international journal of robotics research  
Motion can be described in several alternative representations, including joint configuration or end-effector spaces, but also more complex topology-based representations that imply a change of Voronoi  ...  This allows for consistent combination of multiple representations (e.g. across task, end-effector and joint space).  ...  This is in strong contrast to thinking of y * 0:T as a lower-dimensional task space like an end-effector space.  ... 
doi:10.1177/0278364913482017 fatcat:pme4hlj5tfbkxk4as4cpkyyciy

The neural dynamics of goal-directed arm movements: A developmental perspective

Stephan K. U. Zibner, Jan Tekulve, Gregor Schoner
2015 2015 Joint IEEE International Conference on Development and Learning and Epigenetic Robotics (ICDL-EpiRob)  
generation of the time courses of virtual trajectories of the hand in Cartesian space, and their transformation into virtual joint trajectories and muscle forces.  ...  to joint space, are decalibrated to reflect earlier stages of development.  ...  This work reflects only the authors' views; the EC is not liable for any use that may be made of the information contained herein.  ... 
doi:10.1109/devlrn.2015.7346134 dblp:conf/icdl-epirob/ZibnerTS15a fatcat:m2ntqat5lng4jkyd5glrzzmgly

Visually guided movements: learning with modular neural maps in robotics

Jean-Luc Buessler, Jean-Philippe Urban
1998 Neural Networks  
Therefore, we address this problem starting with the architecture of the system. We illustrate this approach using a robotic application: the visual servoing of the arm's end-effector.  ...  This paper proposes the idea that biological learning can take advantage of the structures of the modules and the nature of modular decomposition.  ...  An output connected to the input of another module constitutes an internal information channel. The data of this channel are named internal representation.  ... 
doi:10.1016/s0893-6080(98)00050-1 pmid:12662757 fatcat:kgo65xjh2fdypks3slbmck7cky

Efficient learning of constraints and generic null space policies

Leopoldo Armesto, Jorren Bosga, Vladimir Ivan, Sethu Vijayakumar
2017 2017 IEEE International Conference on Robotics and Automation (ICRA)  
A large class of motions can be decomposed into a movement task and null-space policy subject to a set of constraints.  ...  This novel formulation of the problem allows the constraint learning method to be coupled with the policy learning method to improve policy learning accuracy, which enables us to learn more complex motions  ...  of the end-effector, and p ee is a vector that represents the position of the end-effector.  ... 
doi:10.1109/icra.2017.7989181 dblp:conf/icra/ArmestoBIV17 fatcat:w47plxmhbzevzgqu6bx275wowe

Adaptive nonparametric kinematic modeling of concentric tube robots

Georgios Fagogenis, Christos Bergeles, Pierre E. Dupont
2016 2016 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)  
We evaluated our approach on data gathered from a three-tube robot, and report high accuracy across the robot's configuration space.  ...  The model comprises an ensemble of linear models, each of which locally approximates the original complex kinematic relation.  ...  Hence, for an n-tube robot, a point in the configuration space is defined as: x = [α i1 , d i1 , ..., α n1 , d n1 ] T ∈ R 2n−2 . The task space comprises all feasible end-effector (tip) poses.  ... 
doi:10.1109/iros.2016.7759636 pmid:28717555 pmcid:PMC5510657 dblp:conf/iros/FagogenisBD16 fatcat:m5kbgltspfgzjhm6cnx5of6k4i
« Previous Showing results 1 — 15 out of 6,130 results