Abstract
We detail an approach to the autonomous acquisition of hierarchical perception-action competences in which capabilities are bootstrapped using an information-based saliency measure.
Our principle aim is hence to accelerate learning in embodied autonomous agents by aggregating novel motor capabilities and their corresponding perceptual representations using a subsumption-based strategy. The method seeks to allocate affordance parameterizations according to the current (possibly autonomously-determined) learning goal in a manner that eliminates redundant percept-motor context, thereby obtaining maximal parametric efficiency.
Experimental results within a simulated environment indicate that doing so reduces the complexity of a multistage perception-action learning problem by several orders of magnitude.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Brooks, R.A.: Intelligence without representation. Artificial Intelligence 47, 139–159 (1991)
Brooks, R.A.: New approaches to robotics. Science 253, 1227–1232 (1991)
Brooks, R.A.: A robust layered control system for a mobile robot, pp. 2–27 (1990)
Castellano, G., Fanelli, A., Mencar, C.: A neuro-fuzzy network to generate human-understandable knowledge from data. Cognitive Systems Research 3, 125–144 (2002)
Fodor, J., Pylyshyn, Z.: Connectionism and cognitive architecture: A critical analysis. Cognition 28, 3–71 (1988)
Giles, L., Omlin, C.W.: Extraction, insertion and refinement of symbolic rules in dynamically driven recurrent neural networks. Connection Science 5, 307–337 (1993)
Kraetzschmar, G.K., Sablatnoeg, S., Enderle, S., Palm, G.: Application of neurosymbolic integration for environment modelling in mobile robots. Hybrid Neural Systems (2000)
Granlund, G.: Organization of architectures for cognitive vision systems. In: Proceedings of Workshop on Cognitive Vision, Schloss Dagstuhl, Germany (October 2003)
Harnad, S.: The symbol grounding problem. Physica D 42, 335–346 (1990)
Kadir, T., Brady, M.: Scale saliency: A novel approach to salient feature and scale selection (2003)
Kaelbling, L.P.: Hierarchical learning in stochastic domains: Preliminary results. In: International Conference on Machine Learning, pp. 167–173 (1993)
Markman, A., Dietrich, E.: Extending the classical view of representation. Trends Cognit. Sci. 4, 470–475 (1991)
McGrenere, J., Ho, W.: Affordances: Clarifying and evolving a concept. In: Proceedings of Graphics Interface 2000, Montreal, Canada, pp. 179–186 (2000)
Sharkey, N., Ziemke, T.: Mechanistic vs. phenomenal embodiment: Can robot embodiment lead to Strong AI? Cognitive Systems Research 2(4), 251–262 (2001)
Spall, J.C.: Introduction to Stochastic Search and Optimization. John Wiley & Sons, Inc., New York (2003)
Steedman, M.: Formalizing affordance. In: 24th Annual Meeting of the Cognitive Science Society, Fairfax VA, pp. 834–839 (2002)
Stoytchev, A.: Toward learning the binding affordances of objects: A behavior-grounded approach. In: Proceedings of AAAI Symposium on Developmental Robotics, Stanford University, March 21-23, 2005, pp. 17–22 (2005)
Sun, R.: Symbol grounding: a new look at an old idea. Philosophical Psychology 13, 149–172 (2000)
Sun, R., Merrill, E., Peterson, T.: From implicit skills to explicit knowledge: a bottom-up model of skill learning. Cognitive Science 25(2), 203–244 (2001)
Tano, S., Futamura, D., Uemura, Y.: Efficient learning by symbol emergence in multilayer network and agent collaboration. In: The Ninth IEEE International Conference on FUZZ IEEE 2000, pp. 1056–1061 (2000)
Ueno, A., Takeda, H., Nishida, T.: Learning of the way of abstraction in real robots. In: Proceedings of IEEE International Conference on IEEE SMC 1999, vol. 2, pp. 746–751 (1999)
Wermter, S.: Knowledge extraction from transducer neural networks. Applied Intelligence: The International Journal of Artificial Intelligence, Neural Networks and Complex Problem-Solving Techniques 12, 27–42 (2000)
Windridge, D., Kittler, J.: Open-ended inference of relational representations in the cospal perception-action architecture. In: The 5th International Conference on Computer Vision Systems (ICVS 2007), Bielefeld, March 21–24 (2007), http://biecoll.ub.uni-bielefeld.de/volltexte/2007/89 , doi:10.2390/biecoll-icvs2007-172
Windridge, D., Kittler, J.: Epistemic constraints on autonomous symbolic representation in natural and artificial agents. In: Smolinski, T.G., Milanova, M.G., Hassanien, A.E. (eds.) Applications of Computational Intelligence in Biology: Current Trends and Open Problems. Studies in Computational Intelligence (SCI), vol. 1. Springer, Heidelberg (2008)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2008 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Windridge, D., Shevchenko, M., Kittler, J. (2008). An Entropy-Based Approach to the Hierarchical Acquisition of Perception-Action Capabilities. In: Caputo, B., Vincze, M. (eds) Cognitive Vision. ICVW 2008. Lecture Notes in Computer Science, vol 5329. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-92781-5_7
Download citation
DOI: https://doi.org/10.1007/978-3-540-92781-5_7
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-92780-8
Online ISBN: 978-3-540-92781-5
eBook Packages: Computer ScienceComputer Science (R0)