Deferred Restructuring of Experience in Autonomous Machines

Informazioni relative al progetto

DREAM

ID dell’accordo di sovvenzione: 640891

DOI 10.3030/640891

Progetto chiuso

Data della firma CE 24 Novembre 2014

Data di avvio 1 Gennaio 2015

Data di completamento 31 Dicembre 2018

Finanziato da

• EXCELLENT SCIENCE - Future and Emerging Technologies (FET)

Costo totale € 2 784 240,94

Contributo UE € 2 730 241,94

Coordinato da

SORBONNE UNIVERSITE
France

Risultati finali

Documents, reports (20)

Publication 1 WP4

First publication on WP4 results.

Publication 3 WP4

Final publication on WP4 results.

Publication 3 WP2

Third publication on WP2 results.

Publication 2 WP3

Second publication on WP3 results.

Data management plan, v1

release of the data management plan, version 1.

Publication 2 WP1

Second publication on WP1 results.

Publication 1 WP3

First publication on WP3 results.

Data management plan, v3

Release of the Data Management plan, version 3.

Publication 1 WP1

Publication on WP1 first results.

Publication 2 WP2

Second publication on WP2 results.

Publication 3 WP1

Third publication on WP1 results.

Publication 3 WP3

Final publication on WP3 results.

Experimental protocol

Experimental protocol to be used in the validation tests for the whole system.

Publication 1 WP5

First publication in between neuroscience and DREAM principles.

Publication WP6

Publication on the whole integrated DREAM architecture.

Publication 1 WP2

Publication on WP2 first results.

Publication 2 WP4

Second publication on WP4 results.

Publication 3 WP5

Final pubication on neuroscience and DREAM principles

Data management plan, v2

Release of the Data Management Plan, version 2.

Publication 2 WP5

Second publication on WP5 results.

Other (6)

DREAM Module WP2

First release of the DREAM module implementing the work done in the WP2

DREAM v1

First version of the DREAM core implementation on top of the CAFER framework.

CAFER Release

first release of the software framework that will support the implementation of the cognitive architecture and its different modules.

DREAM module WP1

First release of the DREAM module implementing the work done in WP1

DREAM Module WP4

First release of the DREAM module implementing the work done in the WP4.

DREAM Module WP3

First release of the DREAM module implementing the work done in WP3.

Open Research Data Pilot (1)

Open Research Data Pilot

Project signed before February 2016 Just to bypasss the error message during the amendement preparation proces.

Websites, patent fillings, videos etc. (1)

DREAM web site

Web site of the project, first release.

Demonstrators, pilots, prototypes (1)

Complete DREAM dem.

Complete DREAM architecture implemented with CAFER on top of the robot chosen for the demonstration.

Pubblicazioni

Peer reviewed articles (15)

Perceptual Generalization and Context in a Network Memory Inspired Long-Term Memory for Artificial Cognition

Autori: Richard J. Duro, Jose A. Becerra, Juan Monroy, Francisco Bellas
Pubblicato in: International Journal of Neural Systems, 2018, Pagina/e 1850053, ISSN 0129-0657
Editore: World Scientific Publishing Co
DOI: 10.1142/S0129065718500533

Evolutionary Robotics: What, Why, and Where to

Autori: Stephane Doncieux, Nicolas Bredeche, Jean-Baptiste Mouret, Agoston E. (Gusz) Eiben
Pubblicato in: Frontiers in Robotics and AI, Numero 2, 2015, ISSN 2296-9144
Editore: Frontiers
DOI: 10.3389/frobt.2015.00004

Many regression algorithms, one unified model: A review

Autori: Freek Stulp, Olivier Sigaud
Pubblicato in: Neural Networks, Numero 69, 2015, Pagina/e 60-79, ISSN 0893-6080
Editore: Pergamon Press Ltd.
DOI: 10.1016/j.neunet.2015.05.005

Contextual modulation of value signals in reward and punishment learning

Autori: Stefano Palminteri, Mehdi Khamassi, Mateus Joffily, Giorgio Coricelli
Pubblicato in: Nature Communications, Numero 6, 2015, Pagina/e 8096, ISSN 2041-1723
Editore: Nature Publishing Group
DOI: 10.1038/ncomms9096

An evolutionary approach to constrained sampling optimization problems

Autori: P. Caama?o, G. Varela, R.J. Duro
Pubblicato in: Applied Soft Computing, Numero 51, 2017, Pagina/e 266-279, ISSN 1568-4946
Editore: Elsevier BV
DOI: 10.1016/j.asoc.2016.12.002

Motivational engine with autonomous sub-goal identification for the Multilevel Darwinist Brain

Autori: R. Salgado, A. Prieto, F. Bellas, L. Calvo-Varela, R.J. Duro
Pubblicato in: Biologically Inspired Cognitive Architectures, Numero 17, 2016, Pagina/e 1-11, ISSN 2212-683X
Editore: Elsevier BV
DOI: 10.1016/j.bica.2016.07.003

Real-time optimization of dynamic problems through distributed Embodied Evolution

Autori: Abraham Prieto, F. Bellas, P. Trueba, R.J. Duro
Pubblicato in: Integrated Computer-Aided Engineering, Numero 23/3, 2016, Pagina/e 237-253, ISSN 1069-2509
Editore: IOS Press
DOI: 10.3233/ICA-160522

Introducing Synaptic Delays in the NEAT Algorithm to Improve Modelling in Cognitive Robotics

Autori: P. Caama?o, R. Salgado, F. Bellas, R. J. Duro
Pubblicato in: Neural Processing Letters, Numero 43/2, 2016, Pagina/e 479-504, ISSN 1370-4621
Editore: Kluwer Academic Publishers
DOI: 10.1007/s11063-015-9426-5

Quantifying Selection Pressure

Autori: Evert Haasdijk, Jacqueline Heinerman
Pubblicato in: Evolutionary Computation, 2017, ISSN 1063-6560
Editore: MIT Press
DOI: 10.1162/EVCO_a_00207

Unsupervised identification and recognition of situations for high-dimensional sensori-motor streams

Autori: Jacqueline Heinerman, Evert Haasdijk, A.E. Eiben
Pubblicato in: Neurocomputing, 2017, ISSN 0925-2312
Editore: Elsevier BV
DOI: 10.1016/j.neucom.2017.02.090

To Cooperate or Not to Cooperate: Why Behavioural Mechanisms Matter

Autori: Arthur Bernard, Jean-Baptiste André, Nicolas Bredeche
Pubblicato in: PLOS Computational Biology, Numero 12/5, 2016, Pagina/e e1004886, ISSN 1553-7358
Editore: -
DOI: 10.1371/journal.pcbi.1004886

Bootstrapping Q-Learning for Robotics from Neuro-Evolution Results

Autori: Matthieu Zimmer, Stephane Doncieux
Pubblicato in: IEEE Transactions on Cognitive and Developmental Systems, 2017, Pagina/e 1-1, ISSN 2379-8920
Editore: IEEE
DOI: 10.1109/TCDS.2016.2628817

Towards Deep Developmental Learning

Autori: Olivier Sigaud, Alain Droniou
Pubblicato in: IEEE Transactions on Cognitive and Developmental Systems, Numero 8/2, 2016, Pagina/e 99-114, ISSN 2379-8920
Editore: IEEE
DOI: 10.1109/TAMD.2015.2496248

Behavioral Specialization in Embodied Evolutionary Robotics: Why So Difficult?

Autori: Jean-Marc Montanier, Simon Carrignon, Nicolas Bredeche
Pubblicato in: Frontiers in Robotics and AI, Numero 3, 2016, ISSN 2296-9144
Editore: -
DOI: 10.3389/frobt.2016.00038

When and where to transfer for Bayesian network parameter learning

Autori: Yun Zhou, Timothy M. Hospedales, Norman Fenton
Pubblicato in: Expert Systems with Applications, Numero 55, 2016, Pagina/e 361-373, ISSN 0957-4174
Editore: Pergamon Press Ltd.
DOI: 10.1016/j.eswa.2016.02.011

Conference proceedings (32)

Facilitating intention prediction for humans by optimizing robot motions

Autori: Freek Stulp, Jonathan Grizou, Baptiste Busch, Manuel Lopes
Pubblicato in: 2015 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2015, Pagina/e 1249-1255, ISBN 978-1-4799-9994-1
Editore: IEEE
DOI: 10.1109/IROS.2015.7353529

Bootstrapping interactions with objects from raw sensorimotor data: A novelty search based approach

Autori: Carlos Maestre, Antoine Cully, Christophe Gonzales, Stephane Doncieux
Pubblicato in: 2015 Joint IEEE International Conference on Development and Learning and Epigenetic Robotics (ICDL-EpiRob), 2015, Pagina/e 7-12, ISBN 978-1-4673-9320-1
Editore: IEEE
DOI: 10.1109/DEVLRN.2015.7346098

Sketch-a-Net that Beats Humans

Autori: Qian Yu, Yongxin Yang, Yi-Zhe Song, Tao Xiang, Timothy Hospedales
Pubblicato in: Procedings of the British Machine Vision Conference 2015, 2015, Pagina/e 7.1-7.12, ISBN 1-901725-53-7
Editore: British Machine Vision Association
DOI: 10.5244/C.29.7

Applying the canonical distributed Embodied Evolution algorithm in a collective indoor navigation task

Autori: P. Trueba, A. Prieto, F. Bellas, R.J. Duro
Pubblicato in: 2015 International Joint Conference on Neural Networks (IJCNN), 2015, Pagina/e 1-8, ISBN 978-1-4799-1960-4
Editore: IEEE
DOI: 10.1109/IJCNN.2015.7280807

Evolution, Individual Learning, and Social Learning in a Swarm of Real Robots

Autori: Jacqueline Heinerman, Massimiliano Rango, A.E. Eiben
Pubblicato in: 2015 IEEE Symposium Series on Computational Intelligence, 2015, Pagina/e 1055-1062, ISBN 978-1-4799-7560-0
Editore: IEEE
DOI: 10.1109/SSCI.2015.152

Three-fold Adaptivity in Groups of Robots - The Effect of Social Learning

Autori: Jacqueline Heinerman, Dexter Drupsteen, A.E. Eiben
Pubblicato in: Proceedings of the 2015 on Genetic and Evolutionary Computation Conference - GECCO '15, 2015, Pagina/e 177-183, ISBN 9781-450334723
Editore: ACM Press
DOI: 10.1145/2739480.2754743

Improving the performance of evolutionary algorithms by soft-constraining their sampling capabilities

Autori: P. Caamano, G. Varela, R.J. Duro
Pubblicato in: 2015 IEEE Congress on Evolutionary Computation (CEC), 2015, Pagina/e 1319-1326, ISBN 978-1-4799-7492-4
Editore: IEEE
DOI: 10.1109/CEC.2015.7257041

When Face Recognition Meets with Deep Learning: an Evaluation of Convolutional Neural Networks for Face Recognition

Autori: G. Hu, Y. Yang, D. Yi, J. Kittler, W. Christmas, S. Z. Li, T. Hospedales
Pubblicato in: ICCV Workshop ChaLearn Looking at People, 2015
Editore: IEEE

Probabilistic Graphical Models Parameter Learning with Transferred Prior and Constraints

Autori: Yun Zhou, Norman Fenton, Timothy M. Hospedales and Martin Neil
Pubblicato in: Uncertainty in Artificial Intelligence, 2015, Pagina/e 972-981
Editore: AUAI Press

Evolution of Cooperation in Evolutionary Robotics: the Tradeoff between Evolvability and Efficiency

Autori: Arthur Bernard, Jean-Baptiste André, Nicolas Bredeche
Pubblicato in: 07/20/2015-07/24/2015, 2015, Pagina/e 495-502, ISBN 9780-262330275
Editore: The MIT Press
DOI: 10.7551/978-0-262-33027-5-ch087

Visual saliency-based babbling of unknown dynamic environments (extended abstract)

Autori: Legoff, L. and Maestre, C. and Doncieux, S.
Pubblicato in: Learning Object Affordances Workshop, IROS 2015, 2015
Editore: Jamone, L. et al.

Learning a high diversity of object manipulations though an evolutionary-based babbling (extended abstract)

Autori: Ecarlat, P. and Cully, A. and Maestre, C. and Doncieux, S.
Pubblicato in: Learning Object Affordances Workshop, IROS 2015, 2015
Editore: Jamone, L. et al.

Multivariate Regression on the Grassmannian for Predicting Novel Domains

Autori: Yongxin Yang, Timothy M. Hospedales
Pubblicato in: 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2016, Pagina/e 5071-5080, ISBN 978-1-4673-8851-1
Editore: IEEE
DOI: 10.1109/CVPR.2016.548

Deep Multi-task Representation Learning: A Tensor Factorisation Approach

Autori: YANG, Yongxin et HOSPEDALES, Timothy
Pubblicato in: Proceedings of ICLR, 2017
Editore: arxiv

Tensor Based Knowledge Transfer Across Skill Categories for Robot Control

Autori: C. Zhao, T. Hospedales, F. Stulp, O. Sigaud
Pubblicato in: Proceedings of IJCAI 2017, 2017
Editore: AAAI Press

Multilevel Darwinist Brain: Context Nodes in a Network Memory Inspired Long Term Memory

Autori: Richard J. Duro, Jose A. Becerra, Juan Monroy, Luis Calvo
Pubblicato in: 2017, Pagina/e 22-31
Editore: Springer International Publishing
DOI: 10.1007/978-3-319-59740-9_3

Embodied Evolution versus Cooperative Coevolution in Multi-Robot Optimization: a practical comparison

Autori: P. Trueba, A. Prieto, F. Bellas
Pubblicato in: Proceedings GECCO 2017, 2017
Editore: ACM

Motivational Engine for Cognitive Robotics in Non-Static Tasks

Autori: Rodrigo Salgado, Abraham Prieto, Francisco Bellas, Richard J. Duro
Pubblicato in: Lecture Notes in Computer Science, Numero 10337, 2017, Pagina/e 32-42
Editore: Springer International Publishing
DOI: 10.1007/978-3-319-59740-9_4

The Robobo Project: Bringing Educational Robotics Closer to Real-World Applications

Autori: F. Bellas, M. Naya, G. Varela, L. Llamas, A. Prieto, J.C. Becerra, M. Bautista, A. Faiña, R.J. Duro
Pubblicato in: Proceedings 8th International Conference on Robotics in Education (RIE 2017), 2017
Editore: Springer

Improving extrinsically motivated developmental robots through intrinsic motivations

Autori: Rodrigo Salgado, Abraham Prieto, Francisco Bellas, Richard J. Duro
Pubblicato in: 2016 Joint IEEE International Conference on Development and Learning and Epigenetic Robotics (ICDL-EpiRob), 2016, Pagina/e 154-155, ISBN 978-1-5090-5069-7
Editore: IEEE
DOI: 10.1109/DEVLRN.2016.7846809

Considering Memory Networks in the LTM Structure of the Multilevel Darwinist Brain

Autori: Richard J. Duro, Jos? Antonio Becerra, Juan Monroy, Pilar Caamano
Pubblicato in: Proceedings of the 2016 on Genetic and Evolutionary Computation Conference Companion - GECCO '16 Companion, 2016, Pagina/e 1057-1060, ISBN 9781-450343237
Editore: ACM Press
DOI: 10.1145/2908961.2931674

MotivEn: Motivational engine with sub-goal identification for autonomous robots

Autori: Rodrigo Salgado, Abraham Prieto, Pilar Caamano, Francisco Bellas, Richard J. Duro
Pubblicato in: 2016 IEEE Congress on Evolutionary Computation (CEC), 2016, Pagina/e 4887-4894, ISBN 978-1-5090-0623-6
Editore: IEEE
DOI: 10.1109/CEC.2016.7744417

How Complexity Pervades Specialization in Canonical Embodied Evolution

Autori: Pedro Trueba, Abraham Prieto, Francisco Bellas, Richard J. Duro
Pubblicato in: Proceedings of the Artificial Life Conference 2016, 2016, Pagina/e 123-130, ISBN 978-0-262-33936-0
Editore: MIT Press
DOI: 10.7551/978-0-262-33936-0-ch026

Motivational Engine with Sub-Goal Identification in Neuroevolution based Cognitive Robotics

Autori: Rodrigo Salgado, Abraham Prieto, Pilar Caama?o, Francisco Bellas, Richard J. Duro
Pubblicato in: Lecture Notes in Computer Science, Numero 9648, 2016, Pagina/e 659-670
Editore: Springer International Publishing
DOI: 10.1007/978-3-319-32034-2_55

On-line Evolution of Foraging Behaviour in a Population of Real Robots

Autori: Jacqueline Heinerman, Alessandro Zonta, Evert Haasdijk, A. E. Eiben
Pubblicato in: Applications of Evolutionary Computation, 2016, Pagina/e 198-212
Editore: Springer International Publishing
DOI: 10.1007/978-3-319-31153-1_14

Is Social Learning More Than Parameter Tuning?

Autori: Heinerman, J., Stork, J., Rebolledo, M., Hubert, J., Eiben, A.E., Bartz-Beielstein, T., Haasdijk, E.
Pubblicato in: Proceedings of the 18th annual conference on Genetic and evolutionary computation, 2017
Editore: ACM
DOI: 10.1145/3067695.3076059

Can Social Learning Increase Learning Speed, Performance or Both?

Autori: Heinerman, J., Stork, J., Rebolledo, M., Hubert, J., Eiben, A.E., Bartz-Beielstein, T., Haasdijk, E.
Pubblicato in: Proceedings of the European Conference on Artificial Life, 2017
Editore: -

An iterative algorithm for forward-parameterized skill discovery

Autori: Adrien Matricon, David Filliat, Pierre-Yves Oudeyer
Pubblicato in: 2016 Joint IEEE International Conference on Development and Learning and Epigenetic Robotics (ICDL-EpiRob), 2016, Pagina/e 186-192, ISBN 978-1-5090-5069-7
Editore: IEEE
DOI: 10.1109/DEVLRN.2016.7846816

Segmenting Objects through an Autonomous Agnostic Exploration Conducted by a Robot

Autori: Leni K. Le Goff, Ghanim Mukhtar, Pierre-Henri Le Fur, Stephane Doncieux
Pubblicato in: 2017 First IEEE International Conference on Robotic Computing (IRC), 2017, Pagina/e 284-291, ISBN 978-1-5090-6724-4
Editore: IEEE
DOI: 10.1109/IRC.2017.22

Evolving Specialisation in a Population of Heterogeneous Robots: the Challenge of Bootstrapping and Maintaining Genotypic Polymorphism

Autori: Arthur Bernard, Jean-Baptiste Andr, Nicolas Bredeche
Pubblicato in: Proceedings of the Artificial Life Conference 2016, 2016, Pagina/e 152-159, ISBN 978-0-262-33936-0
Editore: MIT Press
DOI: 10.7551/978-0-262-33936-0-ch032

Gated Neural Networks for Option Pricing: Rationality by Design

Autori: Y. Yang, Y. Zheng, T. M. Hospedales
Pubblicato in: AAAI Conference on Artificial Intelligence, 2017
Editore: -

Sketch Me That Shoe

Autori: Qian Yu, Feng Liu, Yi-Zhe Song, Tao Xiang, Timothy M. Hospedales, Chen-Change Loy
Pubblicato in: The IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2016, Pagina/e 799-807
Editore: Computer Vision Foundation

Book chapters (3)

Autonomous Learning of Procedural Knowledge in an Evolutionary Cognitive Architecture for Robots

Autori: Rodrigo Salgado, Francisco Bellas, Richard J. Duro
Pubblicato in: Applications of Evolutionary Computation, 2015, Pagina/e 807-818, ISBN 978-3-319-16549-3
Editore: Springer International Publishing
DOI: 10.1007/978-3-319-16549-3_65

Embodied Evolution for Collective Indoor Surveillance and Location

Autori: Pedro Trueba, Abraham Prieto, Francisco Bellas, Richard J. Duro
Pubblicato in: Bioinspired Computation in Artificial Systems, 2015, Pagina/e 138-147, ISBN 978-3-319-18833-1
Editore: Springer International Publishing
DOI: 10.1007/978-3-319-18833-1_15

Studying the Coupled Learning of Procedural and Declarative Knowledge in Cognitive Robotics

Autori: Rodrigo Salgado, Francisco Bellas, Richard J. Duro
Pubblicato in: Biomimetic and Biohybrid Systems, 2015, Pagina/e 304-315, ISBN 978-3-319-22979-9
Editore: Springer International Publishing
DOI: 10.1007/978-3-319-22979-9_30