Journal of the Institute of Electronics Engineers of Korea SC (전자공학회논문지SC)

The Institute of Electronics and Information Engineers (대한전자공학회)
권호 표지

Behavioral motivation-based Action Selection Mechanism with Bayesian Affordance Models

베이지안 행동유발성 모델을 이용한 행동동기 기반 행동 선택 메커니즘

  • Lee, Sang-Hyoung (Department of computer Science & Engineering, Hanyang University) ;
  • Suh, Il-Hong (Department of computer Science & Engineering, Hanyang University)
  • Published : 2009.07.25
Download PDF
⟨ Previous Next ⟩

Abstract

A robot must be able to generate various skills to achieve given tasks intelligently and reasonably. The robot must first learn affordances to generate the skills. An affordance is defined as qualities of objects or environments that induce actions. Affordances can be usefully used to generate skills. Most tasks require sequential and goal-oriented behaviors. However, it is usually difficult to accomplish such tasks with affordances alone. To accomplish such tasks, a skill is constructed with an affordance and a soft behavioral motivation switch for reflecting goal-oriented elements. A skill calculates a behavioral motivation as a combination of both presently perceived information and goal-oriented elements. Here, a behavioral motivation is the internal condition that activates a goal-oriented behavior. In addition, a robot must be able to execute sequential behaviors. We construct skill networks by using generated skills that make action selection feasible to accomplish a task. A robot can select sequential and a goal-oriented behaviors using the skill network. For this, we will first propose a method for modeling and learning Bayesian networks that are used to generate affordances. To select sequential and goal-oriented behaviors, we construct skills using affordances and soft behavioral motivation switches. We also propose a method to generate the skill networks using the skills to execute given tasks. Finally, we will propose action-selection-mechanism to select sequential and goal-oriented behaviors using the skill network. To demonstrate the validity of our proposed methods, "Searching-for-a-target-object", "Approaching-a-target-object", "Sniffing-a-target-object", and "Kicking-a-target-object" affordances have been learned with GENIBO (pet robot) based on the human teaching method. Some experiments have also been performed with GENIBO using the skills and the skill networks.

로봇이 지능적이고 합리적으로 임무를 수행하기 위해서는 다양한 솜씨(skill)가 필요하다. 우리는 솜씨를 생성하기 위해 우선 행동유발성(affordance)을 학습한다. 행동유발성은 행동을 유발하게 하는 물체 또는 환경의 성질로써 솜씨를 생성하는데 유용하게 사용될 수 있다. 로봇이 수행하는 대부분의 임무는 순차적이고 목표 지향적인 행동을 필요로 한다. 그러나 행동유발성만을 이용하여 이러한 임무를 수행하는 것은 쉽지 않다. 이를 위해 우리는 행동유발성과 목표 지향적 요소를 반영하기 위한 소프트 행동동기 스위치(soft behavioral motivation switch)를 이용하여 솜씨를 생성한다. 솜씨는 현재 인지된 정보와 목표 지향적 요소를 결합하여 행동동기를 생성한다. 여기서 행동동기는 목표 지향적인 행동을 활성화시키기 위한 내부 상태를 말한다. 또한, 로봇은 임무 수행을 위해 순차적인 행동 선택을 필요로 한다. 우리는 목표 지향적이고 순차적인 행동 선택이 가능하도록 솜씨를 이용하여 솜씨 네트워크(skill network)를 생성한다. 로봇은 솜씨 네트워크를 이용하여 목표 지향적이고 순차적인 행동을 선택할 수 있다. 본 논문에서는 베이지안 네트워크를 이용한 행동유발성 모델링 및 학습 방법, 행동유발성과 소프트 행동동기 스위치를 이용한 솜씨 및 솜씨 네트워크 생성 방법, 마지막으로 솜씨 네트워크를 이용한 목표 지향적 행동 선택 방법을 제안한다. 우리의 방법을 증명하기 위해 제니보(애완 로봇)를 이용한 교시 기반 학습 방법을 통해 "물체 찾기", "물체에 접근하기", "물체의 냄새 맡기", 그리고 "물체를 발로 차기" 행동유발성들을 학습하였다. 또한, 이들을 이용하여 솜씨 및 솜씨 네트워크를 생성하여 제니보에 적용하고 실험하였다.

Keywords

References

  1. P. Perez-Paredes and M. Sanchez-Tornel, Handbook of Research on E-Learning Methodologies for Language Acquisition, Information Science Reference, pp. 3, 2008
  2. J. J. Gibson, The Ecological Approach to Visual Perception, Houghton Mifflin, Boston, 1979
  3. E. Rome, J. Hertzberg, G. Dorffiner, and P. Doherty, 'Towards Affordance-based Robot Control,' in Proc. of Dagstuhl Seminar 062031, June 5-9, 2006 https://doi.org/10.1007/978-3-540-77915-5
  4. T. Kiyoharu, K. Kenta, and I. Katsumi, 'Approach to Artificial Skill from Affordance Theory-Memory and Embodiment,' in Proc. of Journal of the Robotics Society of Japan, vol. 22, no. 7, pp. 892-900, 2004 https://doi.org/10.7210/jrsj.22.892
  5. W. Lidwell, K. Holden, and J. Butler, Universal Principles of Design, Rockport Publishers, pp. 20, 2003
  6. Dweck and S. Carol, 'Motivational processes affecting learning,' in Proc. of American Psychological Association, vol. 41, pp. 1040-1048, 1986 https://doi.org/10.1037/0003-066X.41.10.1040
  7. S. Russell and P. Norvig, Artificial Intelligence: A Modern Approach, Prentice Hall, ch. 14, 2003
  8. D. Heckerman, 'A tutorial on learning with Bayesian network,' Technical report, MS Research Advanced Technology Division, 1995
  9. http://www.dasarobot.com
  10. S. Lee, I. H. Suh, and W. Y. Kwon, 'A Motivation-based Action-Selection-Mechanism involving Reinforcement Learning,' in Proc. of International Journal of Control, Automation, and Systems, vol. 6, no. 6, pp. 904-914, December, 2008
  11. E. de Sevin and D. Thalmann, 'A Motivation Model of Action Selection for Virtual Humans,' in Proc. of Computer Graphics International 2005 (CGI'05), Stony Brook, NY, USA, June 22-24, 2005 https://doi.org/10.1109/CGI.2005.1500419
  12. E. Beaudry, Y. Brosseau, C. Cote, C. Raievsky, D. Letourneau, F. Kabanza, and F. Michaud, 'Reactive Planning in a Motivated Behavioral Architecture,' in Proc. of American Association for Artificial Intelligence, pp.1242-1247, 2005
  13. E. Lazkano, B. Sierra, A. Astigarrage, and J. M. Otzeta, 'On the use of Bayesian networks to develop behaviours for mobile robots,' Robotics and Autonomous Systems, vol. 55, pp. 253-265, 2007 https://doi.org/10.1016/j.robot.2006.08.003
  14. A. Dearden and Y. Demiris, 'Learning Forward Models for Robots,' in Proc. of the International Joint Conference on Artificial Intelligence, pp. 1440-1445, Edinburgh, , 2005
  15. O. Lebeltel, P. Bessiere, J. Diard, and E. Mazer, 'Bayesian Robot Programming,' Autonomous Robots, pp. 49-79, 2004 https://doi.org/10.1023/B:AURO.0000008671.38949.43
  16. L. Montesano, M. Lopes, A. Bernardino, and J. Santos-Victor, 'Modeling Affordances using Bayesian networks,' in Proc. of IEEE/RSJ Int. Conference in Intelligent Robots and Systems, San Diego, USA, October, 2007 https://doi.org/10.1109/IROS.2007.4399511
  17. M. Raubal and R. Moratz, 'A Functional Model for Affordance-Based Agents,' In: E. Rome et al. (Eds.): Affordance-Based Robot Control, LNAI 4760, pp. 91-105, 2008 https://doi.org/10.1007/978-3-540-77915-5_7
  18. A. Stoychev, 'Behavior-Grounded Representation of Tool Affordances,' in Proc. of the 2005 IEEE International Conference on Robotics and Auto- mation, Barcelona, Spain, April, 2005