The Integration of Action and Language in Cognitive Robots
- When?
- Tuesday 31 March 2009, 14:00 to 15:00
- Where?
- University of Plymouth
- Open to:
- Staff, Students
Prof Angelo Cangelosi, Adaptive Behaviour and Cognition Research Group, School of Computing and Mathematics
Abstract:
Recent theoretical and experimental research on action and language processing in humans and animals clearly demonstrates the strict interaction and co-dependence between language and action. These studies on action-language integration have important implication for the design of communication and linguistic capabilities in cognitive systems and robots (Cangelosi et al. 2007; Cangelosi et al., in preparation). Amongst the various approaches to the design communication capabilities in interactive agents, some provide a more integrative vision of language and treat it as an integral part of the whole cognitive system (Cangelosi & Harnad 2000). The agent’s linguistic abilities are strictly dependent on, and grounded in, other behaviours and skills. Such a strict action-language interaction supports the bootstrapping of the agent’s cognitive system, e.g. through the transfer of properties of action knowledge to that of linguistic representations, and vice versa.
We will review a series of cognitive agents and robotics models of action and language learning. These modes share the following properties: (i) symbols are directly grounded in the agents’ own sensorimotor and cognitive abilities and (ii) the communicative/linguistic behavior are acquired (evolutionarily and developmentally) through the interaction of agents in their physical and social environment. For example, an artificial life model studies the evolutionary origins of linguistic categories, such as nouns and verbs, and their grounding in sensorimotor abilities (Cangelosi 2001). Here the techniques of categorical perception and synthetic brain imaging are used to analyze the sensorimotor bases of linguistic structure. In a further model of action and language learning in cognitive robots we show how agents can autonomously evolve grasping and object manipulation skills (Massera, Cangelosi & Nolfi, 2007) and learn to use name of actions to acquire higher-order sensorimotor categories (Cangelosi & Riga, 2006). We will then discuss the extension of this robotic modeling work to experiments with humanoid robots within the EU project “ITALK: Integration and Transfer of Action and Language Knowledge in Robots” (italkproject.org). A roadmap for future research in action and language integration in cognitive robots if finally proposed.
References and links to papers:
Cangelosi A. (2001). Evolution of communication and language using signals, symbols and words. IEEE Transactions on Evolutionary Computation. 5(2), 93-101
http://www.tech.plym.ac.uk/soc/staff/angelo/papers/Cangelosi-IEEE-2001.pdf
Cangelosi A. & Harnad S. (2000). The adaptive advantage of symbolic theft over sensorimotor toil: Grounding language in perceptual categories. Evolution of Communication, 4(1), 117-142
http://www.tech.plym.ac.uk/soc/staff/angelo/papers/cangelosi-harnad-evocom.pdf
Cangelosi A, Riga T (2006). An embodied model for sensorimotor grounding and grounding transfer: Experiments with epigenetic robots, Cognitive Science, 30(4), 673-689
http://www.tech.plym.ac.uk/soc/staff/angelo/papers/cangelosi-riga-cognitivescience2006.pdf
Cangelosi A, Tikhanoff V., Fontanari J.F., Hourdakis E. (2007). Integrating language and cognition: A cognitive robotics approach. IEEE Computational Intelligence Magazine, 2(3), 65-70
Contact Details
Angelo Cangelosi
Adaptive Behaviour and Cognition Research Group
School of Computing and Mathematics
University of Plymouth