Aris Alissandrakis
ABSTRACT
Imitation is a powerful learning tool when humans and robots interact in a social context. A series of experimental runs and a small pilot user study were conducted to evaluate the performance of a system designed for robot imitation. Performance assessments of similarity of imitative behaviours were carried out by machines and by humans: the system was evaluated quantitatively (from a machine-centric perspective) and qualitatively (from a human perspective) in order to study the reconciliation of these views. The experimental results presented here illustrate how the number of exceptions can be used as a performance measure by a robotic or software imitator of an object manipulation behaviour. (In this context, exceptions are events when the optimal displacement and/or rotation that minimize the dissimilarity metrics used to generate a corresponding imitative behaviour cannot be directly achieved in the particular context.) Results of the user study giving similarity judgments on imitative behaviours were used to examine how the quantitative measure of the number of exceptions (from a robot's perspective) corresponds to the qualitative evaluation of similarity (from a human's perspective) for the imitative behaviours generated by the jabberwocky system. Results suggest that there is a good alignment between this quantitive system centered assessment and the more qualitative human-centered assessment of imitative performance.
- A. Alissandrakis, C. L. Nehaniv, K. Dautenhahn, and J. Saunders. Achieving corresponding effects on multiple robotic platforms: Imitating using different effect metrics. In Proc. Third International Symposium on Imitation in Animals and Artifacts -- Hatfield, UK, 12-14 April 2005, pages 10--19. Society for the Study of Artificial Intelligence and Simulation of Behaviour, 2005.Google Scholar
- A. Alissandrakis, C. L. Nehaniv, K. Dautenhahn, and J. Saunders. An approach for programming robots by demonstration to manipulate objects: Considerations on metrics to achieve corresponding effects. In Proc. 6th IEEE International Symposium on Computational Intelligence in Robotics and Automation (CIRA '05), pages 61--66, 2005.Google Scholar
- A. Billard, Y. Epars, S. Calinon, G. Cheng, and S. Schaal. Discovering optimal imitation strategies. Robotics and Autonomous Systems, 47:2-3, 2004.Google ScholarCross Ref
- C. L. Breazeal, D. Buchsbaum, J. Gray, D. Gatenby, and B. Blumberg. Learning from and about others: Towards using imitation to bootstrap the social understanding of others by robots. Artificial Life, 11(1--2):31--62, 2005. Google ScholarDigital Library
- K. Dautenhahn. Getting to know each other -- artificial social intelligence for autonomous robots. Robotics and Autonomous Systems, 16:333--356, 1995.Google ScholarCross Ref
- Y. Demiris and M. Johnson. Distributed, predictive perception of actions: a biologically inspired robotics architecture for imitation and learning. Connection Science Journal, 15(4):231--243, 2003.Google ScholarCross Ref
- R. Dillmann. Teaching and learning of robot tasks via observation of human performance. Journal of Robotics & Autonomous Systems, 47(2--3):109--116, 2004.Google ScholarCross Ref
- Y. Kuniyoshi, M. Inaba, and H. Inoue. Learning by watching: Extracting reusable task knowledge from visual observations of human performance. IEEE Trans. Robot. Automat., 10:799--822, November 1994.Google ScholarCross Ref
- C. L. Nehaniv and K. Dautenhahn. Mapping between dissimilar bodies: Affordances and the algebraic foundations of imitation. In J. Demiris and A. Birk, editors, Proceedings European Workshop on Learning Robots 1998 (EWLR-7), Edinburgh, 20 July 1998, pages 64--72, 1998.Google Scholar
- C. L. Nehaniv and K. Dautenhahn. Like me? - measures of correspondence and imitation. Cybernetics and Systems, 32(1-2):11--51, 2001.Google ScholarCross Ref
- M. N. Nicolescu and M. M. MatariĆ. Learning and interacting in human-robot domains. IEEE Transactions on Systems, Man, and Cybernetics, Part A, 31(5):419--430, 2001. Google ScholarDigital Library
- S. Schaal. Is imitation learning the route to humanoid robots? Trends in Cognitive Sciences, 3(6):233--242, 1999.Google ScholarCross Ref
Index Terms
- Evaluation of robot imitation attempts: comparison of the system's and the human's perspectives
Recommendations
Correspondence Mapping Induced State and Action Metrics for Robotic Imitation
This paper addresses the problem of body mapping in robotic imitation where the demonstrator and imitator may not share the same embodiment [degrees of freedom (DOFs), body morphology, constraints, affordances, and so on]. Body mappings are formalized ...
Whole-body humanoid robot imitation with pose similarity evaluation
Imitation is considered to be a kind of social learning that allows the transfer of information, actions, behaviors, etc. Whereas current robots are unable to perform as many tasks as human, it is a natural way for them to learn by imitations, just as ...
Human–robot interaction review and challenges on task planning and programming
The wide interest of research and industry in the human–robot interaction HRI related topics is proportional to the increased productivity and flexibility of the production lines, as it combines human and robot capabilities. This paper presents a review ...
Reviews
Goran Trajkovski
With the increasing evidence of the importance of learning via imitation in developmental psychology, studies of robot imitation are gaining momentum as a category of research in the interdisciplinary field of human-robot interaction (HRI). Following the research methodologies of the parent discipline of human-computer interaction (HCI), research labeled as HRI has already revitalized "traditional" (heavily artificial intelligence (AI)-based) robotics at a qualitatively new level. The hope now is that the merger of disciplines that HRI relies on will finally produce a man-made artifact that will truly be capable of imitating human behavior, a goal that AI failed to successfully reach. In this paper, the authors present a pilot study in robot imitation. The task is simple (object manipulation), and the study focuses on both quantitative and qualitative analysis of the experiments, respective to the machine-centric and user-centric evaluation. By playing with the effect metrics and introducing exceptions, the authors observe the imitation attempts of a Jabberwocky-based architecture, which, in principle, means that the system has a module on what to imitate and another on how to imitate a human action, and in an iterative fashion it learns how to successfully achieve the action it is attempting to mimic. This paper, as well as most of those included in the proceedings of the HRI 2006 conference, is interesting and informative. The story is told efficiently, and is well illustrated by figures and helpful captions. Online Computing Reviews Service
Access critical reviews of Computing literature here
Become a reviewer for Computing Reviews.
Comments