My research concerns programming robotic agents by demonstration. I'm developing a robotics simulation platform which integrates a physics engine and neural networks to create robotic agents that learn by imitation and through a natural language interface.
Clicking on the links below will bring you to some didactic interactive games I developed to explain artificial intelligence algorithms like neural networks, genetic algorithms and cellular automata.
Conference papers
- Greco A., Riga T. & Cangelosi A. (2003), The acquisition of new categories through grounded
symbols: An extended connectionist model. In O. Kaynak, E. Alpaydin, E. Oja & L. Xu (Eds.), Artificial Neural Networks and Neural Information Processing - ICANN/ICONIP 2003. Berlin: Springer, pp. 773-770 pdf file
Papers in preparation
- Greco A., Cangelosi A. & Riga T.
(submitted), The grounding of symbols through the
experience with the world: A connectionist approach. Intelligent
Systems (in Italian).
Published abstracts
- Riga T. (2003), Symbol grounding
transfer with neural networks: implications for Alife models. 1st
Italian Alife workshop, Cosenza, September 2003 (in Italian) html file
- Greco A., Riga T., Cangelosi A. (2002),
Grounding transfer of symbols in neural networks. 1st Conference of the
Italian Cognitive Science Association, Rovereto, September 2002 (in
Italian) pdf file
- Greco A., Riga T. & Cangelosi A. (2001), The acquisition of new categories in neural networks
through grounded symbols. Proceedings of the Annual Meeting of the
Italian Association for Experimental Psychology, Alghero, September
2001 (in Italian) html file
- Riga T., Greco A. & Cangelosi
A. (2001), Acquiring higher-level categories by sensorimotor grounding of
entry-level features. Proceedings of the 18th Annual Workshop of the
European Society for the Study of Cognitive Systems, Maastricht, May 2001.
