HRI: Cognitive Models and The Theory of Mind Nikoletta Xirakia - - PowerPoint PPT Presentation

hri cognitive models and the theory of mind
SMART_READER_LITE
LIVE PREVIEW

HRI: Cognitive Models and The Theory of Mind Nikoletta Xirakia - - PowerPoint PPT Presentation

MIN-Fakultt Fachbereich Informatik HRI: Cognitive Models and The Theory of Mind Nikoletta Xirakia Universitt Hamburg Fakultt fr Mathematik, Informatik und Naturwissenschaften Fachbereich Informatik Technische Aspekte Multimodaler


slide-1
SLIDE 1

MIN-Fakultät Fachbereich Informatik

HRI: Cognitive Models and The Theory of Mind

Nikoletta Xirakia

Universität Hamburg Fakultät für Mathematik, Informatik und Naturwissenschaften Fachbereich Informatik Technische Aspekte Multimodaler Systeme

  • 28. November 2016
  • N. Xirakia – HRI: Cognitive Models & ToM

1 / 26

slide-2
SLIDE 2

Outline

Introduction Model of ToM Humans’ perception Robots’ perception Conclusion References

  • 1. Introduction
  • 2. Model of ToM
  • 3. Humans’ perception
  • 4. Robots’ perception
  • 5. Conclusion
  • 6. References
  • N. Xirakia – HRI: Cognitive Models & ToM

2 / 26

slide-3
SLIDE 3

Motivation

Introduction Model of ToM Humans’ perception Robots’ perception Conclusion References

Achieving a successful Human-Robot Interaction requires both partners to have sufficient perception of each other’s actions. For Humans:

◮ Appearance of robot ◮ Communication

For Robots:

◮ Attribute mental states (ToM) ◮ Understand humans

[6]

  • N. Xirakia – HRI: Cognitive Models & ToM

3 / 26

slide-4
SLIDE 4

Theory of Mind

Introduction Model of ToM Humans’ perception Robots’ perception Conclusion References

The Theory of Mind (ToM) is the ability to attribute mental states, to self or others: [14]

◮ Beliefs ◮ Intentions and desires ◮ Thoughts ◮ Emotions

Key capability for:

◮ Cognitive development ◮ Social Interaction

[7]

  • N. Xirakia – HRI: Cognitive Models & ToM

4 / 26

slide-5
SLIDE 5

Model of ToM

Introduction Model of ToM Humans’ perception Robots’ perception Conclusion References

Competing views in the context of “belief and desire” reasoning:

◮ Conceptual change (theory-theory)

◮ Set of laws, theories about beliefs and desires ◮ Explain and predict behaviours and desires

◮ Simulation Theory

◮ Representation of others’ mental state ◮ Using own decision-making

◮ The Theory of Mind Mechanism (ToMM)

◮ Generation and representation of multiple beliefs

  • N. Xirakia – HRI: Cognitive Models & ToM

5 / 26

slide-6
SLIDE 6

Model of ToM

Introduction Model of ToM Humans’ perception Robots’ perception Conclusion References

Competing views in the context of “belief and desire” reasoning:

◮ Conceptual change (theory-theory)

◮ Set of laws, theories about beliefs and desires ◮ Explain and predict behaviours and desires

◮ Simulation Theory

◮ Representation of others’ mental state ◮ Using own decision-making

◮ The Theory of Mind Mechanism (ToMM)

◮ Generation and representation of multiple beliefs

  • N. Xirakia – HRI: Cognitive Models & ToM

5 / 26

slide-7
SLIDE 7

Model of ToM

Introduction Model of ToM Humans’ perception Robots’ perception Conclusion References

Competing views in the context of “belief and desire” reasoning:

◮ Conceptual change (theory-theory)

◮ Set of laws, theories about beliefs and desires ◮ Explain and predict behaviours and desires

◮ Simulation Theory

◮ Representation of others’ mental state ◮ Using own decision-making

◮ The Theory of Mind Mechanism (ToMM)

◮ Generation and representation of multiple beliefs

  • N. Xirakia – HRI: Cognitive Models & ToM

5 / 26

slide-8
SLIDE 8

Model of ToM

Introduction Model of ToM Humans’ perception Robots’ perception Conclusion References

Competing views in the context of “belief and desire” reasoning:

◮ Conceptual change (theory-theory)

◮ Set of laws, theories about beliefs and desires ◮ Explain and predict behaviours and desires

◮ Simulation Theory

◮ Representation of others’ mental state ◮ Using own decision-making

◮ The Theory of Mind Mechanism (ToMM)

◮ Generation and representation of multiple beliefs

  • N. Xirakia – HRI: Cognitive Models & ToM

5 / 26

slide-9
SLIDE 9

Model of ToM

Introduction Model of ToM Humans’ perception Robots’ perception Conclusion References

Competing views in the context of “belief and desire” reasoning:

◮ Conceptual change (theory-theory)

◮ Set of laws, theories about beliefs and desires ◮ Explain and predict behaviours and desires

◮ Simulation Theory

◮ Representation of others’ mental state ◮ Using own decision-making

◮ The Theory of Mind Mechanism (ToMM)

◮ Generation and representation of multiple beliefs

  • N. Xirakia – HRI: Cognitive Models & ToM

5 / 26

slide-10
SLIDE 10

Model of ToM

Introduction Model of ToM Humans’ perception Robots’ perception Conclusion References

Competing views in the context of “belief and desire” reasoning:

◮ Conceptual change (theory-theory)

◮ Set of laws, theories about beliefs and desires ◮ Explain and predict behaviours and desires

◮ Simulation Theory

◮ Representation of others’ mental state ◮ Using own decision-making

◮ The Theory of Mind Mechanism (ToMM)

◮ Generation and representation of multiple beliefs

  • N. Xirakia – HRI: Cognitive Models & ToM

5 / 26

slide-11
SLIDE 11

Model of ToM

Introduction Model of ToM Humans’ perception Robots’ perception Conclusion References

Competing views in the context of “belief and desire” reasoning:

◮ Conceptual change (theory-theory)

◮ Set of laws, theories about beliefs and desires ◮ Explain and predict behaviours and desires

◮ Simulation Theory

◮ Representation of others’ mental state ◮ Using own decision-making

◮ The Theory of Mind Mechanism (ToMM)

◮ Generation and representation of multiple beliefs

  • N. Xirakia – HRI: Cognitive Models & ToM

5 / 26

slide-12
SLIDE 12

Model of ToM

Introduction Model of ToM Humans’ perception Robots’ perception Conclusion References

Competing views in the context of “belief and desire” reasoning:

◮ Conceptual change (theory-theory)

◮ Set of laws, theories about beliefs and desires ◮ Explain and predict behaviours and desires

◮ Simulation Theory

◮ Representation of others’ mental state ◮ Using own decision-making

◮ The Theory of Mind Mechanism (ToMM)

◮ Generation and representation of multiple beliefs

  • N. Xirakia – HRI: Cognitive Models & ToM

5 / 26

slide-13
SLIDE 13

True & False beliefs

Introduction Model of ToM Humans’ perception Robots’ perception Conclusion References

Ability to distinct between true and false beliefs:

◮ True-beliefs (TB)

◮ Beliefs which are true in the physical world

◮ False-beliefs (FB)

◮ Beliefs that others may have, but are not actually true

This ability indicates evidence that a person can appreciate the distinction between the mind and the world. [10]

  • N. Xirakia – HRI: Cognitive Models & ToM

6 / 26

slide-14
SLIDE 14

True & False beliefs

Introduction Model of ToM Humans’ perception Robots’ perception Conclusion References

Ability to distinct between true and false beliefs:

◮ True-beliefs (TB)

◮ Beliefs which are true in the physical world

◮ False-beliefs (FB)

◮ Beliefs that others may have, but are not actually true

This ability indicates evidence that a person can appreciate the distinction between the mind and the world. [10]

  • N. Xirakia – HRI: Cognitive Models & ToM

6 / 26

slide-15
SLIDE 15

True & False beliefs

Introduction Model of ToM Humans’ perception Robots’ perception Conclusion References

Ability to distinct between true and false beliefs:

◮ True-beliefs (TB)

◮ Beliefs which are true in the physical world

◮ False-beliefs (FB)

◮ Beliefs that others may have, but are not actually true

This ability indicates evidence that a person can appreciate the distinction between the mind and the world. [10]

  • N. Xirakia – HRI: Cognitive Models & ToM

6 / 26

slide-16
SLIDE 16

True & False beliefs

Introduction Model of ToM Humans’ perception Robots’ perception Conclusion References

Ability to distinct between true and false beliefs:

◮ True-beliefs (TB)

◮ Beliefs which are true in the physical world

◮ False-beliefs (FB)

◮ Beliefs that others may have, but are not actually true

This ability indicates evidence that a person can appreciate the distinction between the mind and the world. [10]

  • N. Xirakia – HRI: Cognitive Models & ToM

6 / 26

slide-17
SLIDE 17

True & False beliefs

Introduction Model of ToM Humans’ perception Robots’ perception Conclusion References

Ability to distinct between true and false beliefs:

◮ True-beliefs (TB)

◮ Beliefs which are true in the physical world

◮ False-beliefs (FB)

◮ Beliefs that others may have, but are not actually true

This ability indicates evidence that a person can appreciate the distinction between the mind and the world. [10]

  • N. Xirakia – HRI: Cognitive Models & ToM

6 / 26

slide-18
SLIDE 18

Humans’ perception

Introduction Model of ToM Humans’ perception Robots’ perception Conclusion References

The form of an object conveys information:

◮ Use and functionalities ◮ Symbolic information to associations related to it ◮ Aesthetics indicate the behaviour of specific parts

  • N. Xirakia – HRI: Cognitive Models & ToM

7 / 26

slide-19
SLIDE 19

Anthropomorphism

Introduction Model of ToM Humans’ perception Robots’ perception Conclusion References

Anthropomorphism, is understood as the attribution of humanlike properties or characteristics to real or imagined non-human agents and objects. [15]

[1]

  • N. Xirakia – HRI: Cognitive Models & ToM

8 / 26

slide-20
SLIDE 20

Anthropomorphism

Introduction Model of ToM Humans’ perception Robots’ perception Conclusion References

Factors affecting humanlikeness of robots:

◮ Embodiment ◮ Verbal communication ◮ Emotions ◮ Gestures

eMuu [2]

  • N. Xirakia – HRI: Cognitive Models & ToM

9 / 26

slide-21
SLIDE 21

Anthropomorphism

Introduction Model of ToM Humans’ perception Robots’ perception Conclusion References

The importance of the robot’s head in HRI:

◮ Non-verbal cues are mediated through the face ◮ Without a face a robot is anonymous ◮ Physiognomy is important for:

◮ Humanlikeness ◮ Knowledge ◮ Sociability

  • N. Xirakia – HRI: Cognitive Models & ToM

10 / 26

slide-22
SLIDE 22

Uncanny Valley

Introduction Model of ToM Humans’ perception Robots’ perception Conclusion References

The terror from something externally alien or un-known, but also strangely familiar.

[16]

◮ Perception of an object with human characteristics ◮ Rationalisation of its actions and appearance ◮ Development of empathy

[13]

  • N. Xirakia – HRI: Cognitive Models & ToM

11 / 26

slide-23
SLIDE 23

The Prisoner’s Dilemma Experiment

Introduction Model of ToM Humans’ perception Robots’ perception Conclusion References

Evoke ToM and monitor brain activation while playing the Prisoners’ Dilemma with different partners. Participants:

◮ Computer (CP) ◮ Functional Robot (FR) ◮ Anthropomorphic robot (AR) ◮ Human confederate (HP)

[12]

  • N. Xirakia – HRI: Cognitive Models & ToM

12 / 26

slide-24
SLIDE 24

Participants

Introduction Model of ToM Humans’ perception Robots’ perception Conclusion References

Anthropomorphic Robot [3] Functional Robot Briefing

  • N. Xirakia – HRI: Cognitive Models & ToM

13 / 26

slide-25
SLIDE 25

Results

Introduction Model of ToM Humans’ perception Robots’ perception Conclusion References

[3]

  • N. Xirakia – HRI: Cognitive Models & ToM

14 / 26

slide-26
SLIDE 26

Robots’ perception

Introduction Model of ToM Humans’ perception Robots’ perception Conclusion References

One of the most important milestones in the development of ToM, is gaining the ability to attribute a false belief task. But can a cognitive agent do that ?

[8]

  • N. Xirakia – HRI: Cognitive Models & ToM

15 / 26

slide-27
SLIDE 27

The Sally-Anne task

Introduction Model of ToM Humans’ perception Robots’ perception Conclusion References

[4]

  • N. Xirakia – HRI: Cognitive Models & ToM

16 / 26

slide-28
SLIDE 28

Cognitive agent model

Introduction Model of ToM Humans’ perception Robots’ perception Conclusion References

ACT-R as core cognitive architecture:

[9]

  • N. Xirakia – HRI: Cognitive Models & ToM

17 / 26

slide-29
SLIDE 29

ToMM

Introduction Model of ToM Humans’ perception Robots’ perception Conclusion References

Stage simulates the Sally-Anne task, which feeds the model with visual information. The agent records:

◮ What happened? ◮ Who saw it happen? ◮ The current location of the marble

False-belief question:

◮ Highest activation - TB answer ◮ Answer incorrect - Sally is not aware ◮ Consideration of possible beliefs

  • N. Xirakia – HRI: Cognitive Models & ToM

18 / 26

slide-30
SLIDE 30

Developmental Mechanisms

Introduction Model of ToM Humans’ perception Robots’ perception Conclusion References

Learning mechanism:

◮ Beginning: Answer based on the highest activation chunk ◮ Learning: Considers if Sally knows about the belief ◮ Maturation parameter:

◮ Gradual ◮ Guideline on the strength of model’s abilities

◮ Selection parameter:

◮ Determines availability of productions (beliefs)

  • N. Xirakia – HRI: Cognitive Models & ToM

19 / 26

slide-31
SLIDE 31

Results

Introduction Model of ToM Humans’ perception Robots’ perception Conclusion References

◮ Model begins to generate multiple beliefs at the age of 2

(approximately)

◮ For selection parameter of 0.5 , does not know to do the

selection

◮ By the age 3.7 years , the selection parameter is up to 0.8 ◮ By the age 5.7 years , the selection parameter equals 0.95 ◮ As the selection parameter increases, so does the efficacy of

learning

  • N. Xirakia – HRI: Cognitive Models & ToM

20 / 26

slide-32
SLIDE 32

Results

Introduction Model of ToM Humans’ perception Robots’ perception Conclusion References

Children [10] Agent [11]

  • N. Xirakia – HRI: Cognitive Models & ToM

21 / 26

slide-33
SLIDE 33

Conclusion

Introduction Model of ToM Humans’ perception Robots’ perception Conclusion References

◮ Anthropomorphism and embodiment have significant effect on

the human interaction, as partners expect a human like behaviour

◮ Humans evoke stronger emotional responses to embodied

agents, as they prove to be able to physically manipulate the environment

◮ ToM is developed by concurrent learning and maturation and

it’s achievable in cognitive agents

◮ The model proved to be a good match to existing data from

developing children, which enables such data to be used for further research on other aspects of ToM and HRI

  • N. Xirakia – HRI: Cognitive Models & ToM

22 / 26

slide-34
SLIDE 34

References

Introduction Model of ToM Humans’ perception Robots’ perception Conclusion References

1. Zlotowski, J., Proudfoot, D., Yogeeswaran, K., & Bartneck, C. (2015). Anthropomorphism: Opportunities and Challenges in Human-Robot Interaction. International Journal of Social Robotics, 7(3), 347-360 2. Bartneck, C., Okada, M. (2001), eMuu An Emotional Robotic User Interface, Philips User Interface Conference (UI2001), Eindhoven 3.

  • F. Hegel, S. Krach, T. Kircher, B. Wrede and G. Sagerer, "Theory of Mind (ToM) on robots: A functional

neuroimaging study,"2008 3rd ACM/IEEE International Conference on Human-Robot Interaction (HRI), Amsterdam, 2008, pp. 335-342. 4. Baron-Cohen, S., Leslie, A. M., Frith, U. (1985). Does the autistic child have a "theory of mind"?. Cognition 21:37-46, p.42 5. Zlotowski, J., Strasser, E., & Bartneck, C. (2014). Dimensions of Anthropomorphism - From Humanness to Humanlikeness. Proceedings of the ACM / IEEE International Conference on Human-Robot Interaction, Bielefeld pp. 66-73. | DOI: 10.1145/2559636.2559679

  • N. Xirakia – HRI: Cognitive Models & ToM

23 / 26

slide-35
SLIDE 35

References

Introduction Model of ToM Humans’ perception Robots’ perception Conclusion References

6. http://www.redorbit.com/news/technology/1112892978/robotics-perception-key-to-interaction-070913/ 7. http://theautismblog.seattlechildrens.org/autism-theory-mind/ 8. https://laughingsquid.com/2nd-annual-robot-film-festival-in-new-york-city/ 9. https://www.cs.umd.edu/class/fall2002/cmsc838s/tichi/actr.html 10. Wellman, H. W., Cross, D., & Watson, J. (2001). Meta- analysis of theory-of-mind development: The truth about false belief. Child Development, 72(3), 655-684 11.

  • L. M. Hiatt and J. G. Trafton. A cognitive model of theory of mind. In Proceedings of ICCM, 2010.

12. https://www.youtube.com/watch?v=t9Lo2fgxWHw 13. http://www.animatorisland.com/the-uncanny-valley/ 14. Dennett, D.C., 1980. The milk of human intentionality (commentary on Searle). Behav. Brain Sci. 3, 428?430. 15.

  • N. Epley, A. Waytz, and J. T. Cacioppo. On seeing human: A three-factor theory of anthropomorphism.

Psychological Review, 114(4):864?886, Oct. 2007. 16.

  • S. Freud (1953). ?The Uncanny?. The Standard Edition of the Complete Psychological Works of Sigmund

Freud, ed & trs. James Strachey, vol. XVII (London: Hogarth), pp. 219-252.

  • N. Xirakia – HRI: Cognitive Models & ToM

24 / 26

slide-36
SLIDE 36

Further reading

Introduction Model of ToM Humans’ perception Robots’ perception Conclusion References

1. Sandoval, E. B., Brandstetter, J., Yacil, U., & Bartneck, C. (2016). Can a Robot Bribe a Human? The Measurement of the Dark Side of Reciprocity in Human Robot Interaction. Proceedings of the 11th ACM/IEEE International Conference on Human-Robot Interaction, Christchurch pp. 117 - 124. | DOI: 10.1109/HRI.2016.7451742 2. Zlotowski, J., Sumioka, H., Nishio, S., Glas, D., Bartneck, C., & Ishiguro, H. (2015). Persistence of the Uncanny Valley: the Influence of Repeated Interactions and a Robot?s Attitude on Its Perception. Frontiers in Cognitive Science, 6(883). | DOI: 10.3389/fpsyg.2015.00883 3. Sandoval, E. B., Brandstetter, J., Obaid, M., & Bartneck, C. (2015). Reciprocity in Human Robot Interaction ? A Quantitative Approach Through The Prisoner?s Dilemma And The Ultimatum Game. International Journal on Social Robotics 8(2), pp 303-317. | DOI: 10.1007/s12369-015-0323-x 4. Bartneck, C., Hoek, M. v. d., Mubin, O., & Mahmud, A. A. (2007). ?Daisy, Daisy, Give me your answer do!? - Switching off a robot. Proceedings of the 2nd ACM/IEEE International Conference on Human-Robot Interaction, Washington DC pp. 217 - 222.

  • N. Xirakia – HRI: Cognitive Models & ToM

25 / 26

slide-37
SLIDE 37

Thank you!

Introduction Model of ToM Humans’ perception Robots’ perception Conclusion References

14

14https://www.zazzle.com.au/thank+you+robot+craft+supplies

  • N. Xirakia – HRI: Cognitive Models & ToM

26 / 26