The IM-CLeVeR Project: Intrinsically Motivated Cumulative Learning Versatile Robots

Gianluca Baldassare; Marco Mirolli; Francesco Mannella; Daniele Caligiore; Elisabetha Visalberghi; Francesco Natale; Valentina Truppa; Gloria Sabbatini; Eugenio Guglielmelli; Flavio Keller; Domenico Campolo; Peter Redgrave; Kevin Gurney; Tom Stafiord; Jochen Triesch; Cornelius Weber; Constantin Rothkopf; Ulrich Nehmzow; Joan Condell; Mia Siddique; Mark Lee; Martin Hülse; Juergen Schmidhuber; Faustino Gomez; Alexander Foester; Julian Togelius; Andrew Barto

The IM-CLeVeR Project: Intrinsically Motivated Cumulative Learning Versatile Robots

Gianluca Baldassare, Marco Mirolli, Francesco Mannella, Daniele Caligiore, Elisabetha Visalberghi, Francesco Natale, Valentina Truppa, Gloria Sabbatini, Eugenio Guglielmelli, Flavio Keller, Domenico Campolo, Peter Redgrave, Kevin Gurney, Tom Stafiord, Jochen Triesch, Cornelius Weber, Constantin Rothkopf, Ulrich Nehmzow, Joan Condell, Mia SiddiqueMark Lee, Martin Hülse, Juergen Schmidhuber, Faustino Gomez, Alexander Foester, Julian Togelius, Andrew Barto

Department of Computer Science

Research output: Chapter in Book/Report/Conference proceeding › Conference Proceeding (Non-Journal item)

Abstract

This short paper presents the core ideas of the IM-CLeVeR Project. IM-CLeVeR aims at developing a new methodology for designing
robot controllers that can: (a) cumulatively learn new skills through autonomous develop-ment based on intrinsic motivations, and (b) reuse such skills for accomplishing multiple, complex, and externally-assigned tasks. This goal will be pursued by investigating three fundamental issues: (a) the mechanisms of
abstraction of sensorimotor information; (b) the mechanisms underlying intrinsic motiva- tions; (c) hierarchical architectures that per- mit cumulative learning. The study of these issues will be conducted on the basis of empir- ical experiments run with monkeys, children, and human adults, with bio-mimetic models aimed at reproducing and interpreting the re- sults of such experiments, and through the de-sign of innovative machine learning systems. The models, architectures, and algorithms so developed will be validated with experiments
and demonstrators run with the simulated and real iCub humanoid robot.

Original language	English
Title of host publication	Proceedings of the 9th International Conference on Epigenetic Robotics
Subtitle of host publication	Modeling Cognitive Development in Robotic Systems
Editors	Lola Cañamero, Pierre-Yves Oudeyer, Christian Balkenius
Publisher	Lund University Publications
Pages	189-190
Number of pages	2
Publication status	Published - 01 Dec 2009
Event	Proceedings of the Ninth International Conference on Epigenetic Robotics: Modeling Cognitive Development in Robotic Systems - Venice, Italy Duration: 12 Nov 2009 → 14 Nov 2009

Conference

Conference	Proceedings of the Ninth International Conference on Epigenetic Robotics: Modeling Cognitive Development in Robotic Systems
Country/Territory	Italy
City	Venice
Period	12 Nov 2009 → 14 Nov 2009

Access to Document

http://homepages.feis.herts.ac.uk/~epirob09/

Permanent link

Persistent link

Cite this

Baldassare, G., Mirolli, M., Mannella, F., Caligiore, D., Visalberghi, E., Natale, F., Truppa, V., Sabbatini, G., Guglielmelli, E., Keller, F., Campolo, D., Redgrave, P., Gurney, K., Stafiord, T., Triesch, J., Weber, C., Rothkopf, C., Nehmzow, U., Condell, J., ... Barto, A. (2009). The IM-CLeVeR Project: Intrinsically Motivated Cumulative Learning Versatile Robots. In L. Cañamero, P.-Y. Oudeyer, & C. Balkenius (Eds.), Proceedings of the 9th International Conference on Epigenetic Robotics: Modeling Cognitive Development in Robotic Systems (pp. 189-190). Lund University Publications. http://homepages.feis.herts.ac.uk/~epirob09/

Baldassare, Gianluca ; Mirolli, Marco ; Mannella, Francesco et al. / The IM-CLeVeR Project : Intrinsically Motivated Cumulative Learning Versatile Robots. Proceedings of the 9th International Conference on Epigenetic Robotics: Modeling Cognitive Development in Robotic Systems. editor / Lola Cañamero ; Pierre-Yves Oudeyer ; Christian Balkenius. Lund University Publications, 2009. pp. 189-190

@inproceedings{59b52bbd00d3484daaa3d5fad92057e1,

title = "The IM-CLeVeR Project: Intrinsically Motivated Cumulative Learning Versatile Robots",

abstract = "This short paper presents the core ideas of the IM-CLeVeR Project. IM-CLeVeR aims at developing a new methodology for designingrobot controllers that can: (a) cumulatively learn new skills through autonomous develop-ment based on intrinsic motivations, and (b) reuse such skills for accomplishing multiple, complex, and externally-assigned tasks. This goal will be pursued by investigating three fundamental issues: (a) the mechanisms ofabstraction of sensorimotor information; (b) the mechanisms underlying intrinsic motiva- tions; (c) hierarchical architectures that per- mit cumulative learning. The study of these issues will be conducted on the basis of empir- ical experiments run with monkeys, children, and human adults, with bio-mimetic models aimed at reproducing and interpreting the re- sults of such experiments, and through the de-sign of innovative machine learning systems. The models, architectures, and algorithms so developed will be validated with experimentsand demonstrators run with the simulated and real iCub humanoid robot.",

author = "Gianluca Baldassare and Marco Mirolli and Francesco Mannella and Daniele Caligiore and Elisabetha Visalberghi and Francesco Natale and Valentina Truppa and Gloria Sabbatini and Eugenio Guglielmelli and Flavio Keller and Domenico Campolo and Peter Redgrave and Kevin Gurney and Tom Stafiord and Jochen Triesch and Cornelius Weber and Constantin Rothkopf and Ulrich Nehmzow and Joan Condell and Mia Siddique and Mark Lee and Martin H{\"u}lse and Juergen Schmidhuber and Faustino Gomez and Alexander Foester and Julian Togelius and Andrew Barto",

note = "Baldassare, G. et al.: The IM-CLeVeR Project: Intrinsically Motivated Cumulative Learning Versatile Robots. In: Canamero, L., Oudeyer, P.-Y., Balkenius, C. (Eds) Proc. of the 9th Int. Conf. on Epigenetic Robotics: Modeling Cognitive Development in Robotic Systems. Lund University Cognitive Studies, 146, pp. 189 - 190, 2009. Sponsorship: EU; Proceedings of the Ninth International Conference on Epigenetic Robotics: Modeling Cognitive Development in Robotic Systems ; Conference date: 12-11-2009 Through 14-11-2009",

year = "2009",

month = dec,

day = "1",

language = "English",

pages = "189--190",

editor = "Lola Ca{\~n}amero and Pierre-Yves Oudeyer and Christian Balkenius",

booktitle = "Proceedings of the 9th International Conference on Epigenetic Robotics",

publisher = "Lund University Publications",

address = "Sweden",

}

Baldassare, G, Mirolli, M, Mannella, F, Caligiore, D, Visalberghi, E, Natale, F, Truppa, V, Sabbatini, G, Guglielmelli, E, Keller, F, Campolo, D, Redgrave, P, Gurney, K, Stafiord, T, Triesch, J, Weber, C, Rothkopf, C, Nehmzow, U, Condell, J, Siddique, M, Lee, M, Hülse, M, Schmidhuber, J, Gomez, F, Foester, A, Togelius, J & Barto, A 2009, The IM-CLeVeR Project: Intrinsically Motivated Cumulative Learning Versatile Robots. in L Cañamero, P-Y Oudeyer & C Balkenius (eds), Proceedings of the 9th International Conference on Epigenetic Robotics: Modeling Cognitive Development in Robotic Systems. Lund University Publications, pp. 189-190, Proceedings of the Ninth International Conference on Epigenetic Robotics: Modeling Cognitive Development in Robotic Systems, Venice, Italy, 12 Nov 2009. <http://homepages.feis.herts.ac.uk/~epirob09/>

The IM-CLeVeR Project: Intrinsically Motivated Cumulative Learning Versatile Robots. / Baldassare, Gianluca; Mirolli, Marco; Mannella, Francesco et al.
Proceedings of the 9th International Conference on Epigenetic Robotics: Modeling Cognitive Development in Robotic Systems. ed. / Lola Cañamero; Pierre-Yves Oudeyer; Christian Balkenius. Lund University Publications, 2009. p. 189-190.

Research output: Chapter in Book/Report/Conference proceeding › Conference Proceeding (Non-Journal item)

TY - GEN

T1 - The IM-CLeVeR Project

T2 - Proceedings of the Ninth International Conference on Epigenetic Robotics: Modeling Cognitive Development in Robotic Systems

AU - Baldassare, Gianluca

AU - Mirolli, Marco

AU - Mannella, Francesco

AU - Caligiore, Daniele

AU - Visalberghi, Elisabetha

AU - Natale, Francesco

AU - Truppa, Valentina

AU - Sabbatini, Gloria

AU - Guglielmelli, Eugenio

AU - Keller, Flavio

AU - Campolo, Domenico

AU - Redgrave, Peter

AU - Gurney, Kevin

AU - Stafiord, Tom

AU - Triesch, Jochen

AU - Weber, Cornelius

AU - Rothkopf, Constantin

AU - Nehmzow, Ulrich

AU - Condell, Joan

AU - Siddique, Mia

AU - Lee, Mark

AU - Hülse, Martin

AU - Schmidhuber, Juergen

AU - Gomez, Faustino

AU - Foester, Alexander

AU - Togelius, Julian

AU - Barto, Andrew

N1 - Baldassare, G. et al.: The IM-CLeVeR Project: Intrinsically Motivated Cumulative Learning Versatile Robots. In: Canamero, L., Oudeyer, P.-Y., Balkenius, C. (Eds) Proc. of the 9th Int. Conf. on Epigenetic Robotics: Modeling Cognitive Development in Robotic Systems. Lund University Cognitive Studies, 146, pp. 189 - 190, 2009. Sponsorship: EU

PY - 2009/12/1

Y1 - 2009/12/1

N2 - This short paper presents the core ideas of the IM-CLeVeR Project. IM-CLeVeR aims at developing a new methodology for designingrobot controllers that can: (a) cumulatively learn new skills through autonomous develop-ment based on intrinsic motivations, and (b) reuse such skills for accomplishing multiple, complex, and externally-assigned tasks. This goal will be pursued by investigating three fundamental issues: (a) the mechanisms ofabstraction of sensorimotor information; (b) the mechanisms underlying intrinsic motiva- tions; (c) hierarchical architectures that per- mit cumulative learning. The study of these issues will be conducted on the basis of empir- ical experiments run with monkeys, children, and human adults, with bio-mimetic models aimed at reproducing and interpreting the re- sults of such experiments, and through the de-sign of innovative machine learning systems. The models, architectures, and algorithms so developed will be validated with experimentsand demonstrators run with the simulated and real iCub humanoid robot.

AB - This short paper presents the core ideas of the IM-CLeVeR Project. IM-CLeVeR aims at developing a new methodology for designingrobot controllers that can: (a) cumulatively learn new skills through autonomous develop-ment based on intrinsic motivations, and (b) reuse such skills for accomplishing multiple, complex, and externally-assigned tasks. This goal will be pursued by investigating three fundamental issues: (a) the mechanisms ofabstraction of sensorimotor information; (b) the mechanisms underlying intrinsic motiva- tions; (c) hierarchical architectures that per- mit cumulative learning. The study of these issues will be conducted on the basis of empir- ical experiments run with monkeys, children, and human adults, with bio-mimetic models aimed at reproducing and interpreting the re- sults of such experiments, and through the de-sign of innovative machine learning systems. The models, architectures, and algorithms so developed will be validated with experimentsand demonstrators run with the simulated and real iCub humanoid robot.

M3 - Conference Proceeding (Non-Journal item)

SP - 189

EP - 190

BT - Proceedings of the 9th International Conference on Epigenetic Robotics

A2 - Cañamero, Lola

A2 - Oudeyer, Pierre-Yves

A2 - Balkenius, Christian

PB - Lund University Publications

Y2 - 12 November 2009 through 14 November 2009

ER -

Baldassare G, Mirolli M, Mannella F, Caligiore D, Visalberghi E, Natale F et al. The IM-CLeVeR Project: Intrinsically Motivated Cumulative Learning Versatile Robots. In Cañamero L, Oudeyer PY, Balkenius C, editors, Proceedings of the 9th International Conference on Epigenetic Robotics: Modeling Cognitive Development in Robotic Systems. Lund University Publications. 2009. p. 189-190

The IM-CLeVeR Project: Intrinsically Motivated Cumulative Learning Versatile Robots

Abstract

Conference

Access to Document

Permanent link

Fingerprint

Cite this