Seminar 200724

1. Walking Your Virtual Dog:
Analysis of Awareness and Proxemics with Simulated
Support Animals in Augmented Reality
ISMAR 2019
Nahal Norouzi, Kangsoo Kim, Myungho Lee, Ryan Schubert, Austin Erickson, Jeremy Bailenson, Gerd Bruder, Greg Welch
Presenter: Seunghyeong Choe
2020. 07. 24

2. Contents
• Overview
• Related Work
• Research Questions
• Experimental Design
• Measures and Hypothesis
• Results
• Discussions

3. Overview of the paper
2
Virtual
Animal
Robotics
Animal
• social companion
• emotional support
• strong social bond
• Human-Animal Interaction
• Emotional interaction
• Co-existence
• Companionship
• Therapy
Application in VR and AR
Understanding of behaviors and characteristics of virtual animals in HMD and their interaction

4. • Psychology: Positive effects of domestic animals [1]
 Blood pressure [2]
 Stress management [3]
 Depression [4]
• Robotics
 Sony Aibo: Comparison of behavior between real dog and Sony Aibo [5]
 Cat toy [6]
 Similar behavioral responses towards both robot and real dog [7]
Related Work
3
[1] S. A. Zinn and A. M. Beck. From the Editors: The humananimal bond and domestication: Through the ages . . . animals in our lives. Animal Frontiers, 4(3):5–6, 2014.
[2] K. Allen. Are pets a healthy pleasure? the influence of pets on blood pressure. Current Directions in Psychological Science, 12(6):236–239, 2003.
[3] J. M. Siegel. Stressful life events and use of physician services among the elderly: the moderating role of pet ownership. Journal of personality and social psychology, 58(6):1081, 1990.
[4] D. M. Collins, S. G. Fitzgerald, N. Sachs-Ericsson, M. Scherer, R. A. Cooper, and M. L. Boninger. Psychosocial well-being and community participation of service dog partners. Disability and Rehabilitation: Assistive Technology, 1(1-2):41–48, 2006.
[5] G. F. Melson, P. H. Kahn Jr, A. M. Beck, B. Friedman, T. Roberts, and E. Garrett. Robots as dogs?: children’s interactions with the robotic dog AIBO and a live australian shepherd. In ACM Human factors in computing systems, pp. 1649–1652, 2005.
[6] K. Thodberg, L. U. Sørensen, P. B. Videbech, P. H. Poulsen, B. Houbak, V. Damgaard, I. Keseler, D. Edwards, and J.W. Christensen. Behavioral responses of nursing home residents to visits from a person with a dog, a robot seal or a toy cat. Anthrozoos, 29(1):107–121, 2016.
[7] A. Kerepesi, E. Kubinyi, G. Jonsson, M. Magnusson, and A. Miklosi. Behavioural comparison of human–animal (dog) and human–robot (aibo) interactions. Behavioural processes, 73(1):92–99, 2006.

5. • VR/AR on human-animal interaction
 The Sims: Pets , Nintendogs
 Animals in a digital classroom increased learning motivation [1]
 Healthy habits in eating [2] and exercise [3]
 Realistic rendering and animation [4, 5, 6, 7]
• Proxemics and Locomotion Behavior
 Proxemics (근접학)
• Managing surrounding space in reaction to other social and nonsocial entities in the space
• Keeping a comfortable distance while walking with other people
 No prior work of the effects of AR animals on proxemics and locomotion behavior
Related Work
4
[1] Z.-H. Chen, C.-Y. Chou, Y.-C. Deng, and T.-W. Chan. Active open learner models as animal companions: Motivating children to learn through interacting with my-pet and our-pet. International Journal of Artificial Intelligence in Education, 17(2):145–167, 2007.
[2] S. J. Ahn, K. Johnsen, J. Moore, S. Brown, M. Biersmith, and C. Ball. Using virtual pets to increase fruit and vegetable consumption in children: A technology-assisted social cognitive theory approach. Cyberpsychology, Behavior, and Social Networking, 19(2):86–92, 2016.
[3] S. J. Ahn, K. Johnsen, T. Robertson, J. Moore, S. Brown, A. Marable, and A. Basu. Using virtual pets to promote physical activity in children: An application of the youth physical activity promotion model. Journal of health communication, 20(7):807–815, 2015.
[4] S. Coros, A. Karpathy, B. Jones, L. Reveret, and M. Van De Panne. Locomotion skills for simulated quadrupeds. In ACM Transactions on Graphics, vol. 30, p. 59, 2011.
[5] X. B. Peng, G. Berseth, and M. Van de Panne. Dynamic terrain traversal skills using reinforcement learning. ACM Transactions on Graphics, 34(4):80, 2015.
[6] L.-Q. Yan, C.-W. Tseng, H. W. Jensen, and R. Ramamoorthi. Physically-accurate fur reflectance: modeling, measurement and rendering. ACM Transactions on Graphics, 34(6):185, 2015.
[7] H. Zhang, S. Starke, T. Komura, and J. Saito. Mode-adaptive neural networks for quadruped motion control. ACM Transactions on Graphics, 37(4):145, 2018.

6. • Existing initial realizations
Related Work
5
Playstation EyePet Microsoft HoloPet Magic Leap Porg

7. Research Questions
6
Q1
Q2
Q3
How does walking an AR dog change its owner’s locomotion behavior
and proxemics in the presence of another person?
How do the AR dog’s awareness and behavior with respect to a person
in the physical environment affect its owner’s perception of the dog
and that person?
How does another person’s apparent ability to see the AR dog affect
the owner’s perception and behavior with respect to the dog
and that person?

8. • Investigate AR dog on,
 Human perception, behavior, locomotion, and proxemics.
• Participants
 21 participants (8 females, 13 males, graduate and undergraduate students)
 14 owned pets
 13 reported experience interacting with virtual pets
 Familiarity
• Computer: 5.38/7
• VR: 4.71/7
• AR: 3.42/7
 4 participants were excluded because of technical issues of HoloLens HMD
Experimental Design
7

9. • AR dog’s appearance and control
 Beagle dog [1, 2] model animated and rendered via the Unity graphics engine
 Following and keeping a distance of 35cm from left or right side of participants
 Set speed to 0.5m/s [3, 4, 5]
 Auditory feedback: panting, barking, or sniffing
 Remotely controlled by a human experimenter using a separate computer
Experimental Design: Material
8
[1] R. Coppinger and L. Coppinger. Dogs: A Startling New Understanding of Canine Origin, Behavior & Evolution. Scribner, 1st ed., 2001.
[2] V. Schwind, K. Leicht, S. J¨ager, K. Wolf, and N. Henze. Is there an uncanny valley of virtual animals? a quantitative and qualitative investigation. International Journal of Human-Computer Studies, 111:49–61,
2018.
[3] G. Goslow, H. Seeherman, C. Taylor, M. McCutchin, and N. Heglund. Electrical activity and relative length changes of dog limb muscles as a function of speed and gait. Journal of Experimental Biology, 94(1):15–42, 1981.
[4] C. T. Rubin and L. E. Lanyon. Limb mechanics as a function of speed and gait: a study of functional strains in the radius and tibia of horse and dog. Journal of experimental biology, 101(1):187–211, 1982.
[5] Swifto: How fast should I walk my dog? . https://swifto.com/blog/howfast-should-i-walk-my-dog, 2019.
6 behavior patterns 4 color patterns

10. • Physical setup
Experimental Design: Material
9
 Preparation Room
• Quiet, isolated
• Give description and instruction
 Human Confederates
• Include co-experimenters
• Perform standardized behaviors
 Interaction Room
• Regular office-like image projected onto the walls
 Walkway
• Measure walking behavior

11. • 2 X 2 mixed factorial design
 Confederate:
• Aware: Saying ”Oh, there’s a dog”
• Unaware: Did not wear a HoloLens
 AR dog:
• Aware: Respond to the collision
• Unaware: Foot pass through the AR dog
Experimental Design: Study Design
10

12. • Phase 1 [Dog Personalization]
 Choosing appearance and naming
• Phase 2 [Play session]
 Wear HoloLens and see the AR dog
 3 minutes to interact and play with verbal commands
Experimental Design: Procedure
11

13. • Phase 3 [Witnessing Collision]
• Phase 4 [Walking without AR Dog]
• Phase 5 [Walking with AR Dog]
• Experience another Phase 3-5 with a
different condition
Experimental Design: Procedure
12

14. • Proxemics and Locomotion Behavior
 Logged head pose tracking information (Phase 4, 5)
Measures and Hypotheses
13
Passing Distance
Walking Speed
Head Rotation
Trajectory Length
Observation Ratio
• Minimum clearance distance to the confederate
• Indicator of participants’ personal space and social presence with other entities
• Average walking speed
• Faster or slower than 0.5m/s: participants’ connection to their dog
• Trajectory of participants’ gaze
• Length of the path
• The time participants looked at their AR dog and divided it by
the total time when the collision happened

15. • AR Animal Perception in a Shared Space
 Questionnaires
Measures and Hypotheses
14
Co-Presence
Godspeed
Perceived Physicality
Affective Attraction
• Basdogan’s Co-Presence questionnaire [1]
• “Anthropomorphism” category of the Godspeed questionnaire
designed by Bartneck et al. [2]
• Trajectory of participants’ gaze
• Length of the path
[1] C. Basdogan, C.-H. Ho, M. A. Srinivasan, and M. Slater. An experimental study on the role of touch in shared virtual environments. ACM Transactions on Computer-Human Interaction, 7(4):443–460, 2000.
[2] C. Bartneck, E. Croft, and D. Kulic. Measurement instruments for the anthropomorphism, animacy, likeability, perceived intelligence, and perceived safety of robots. International Journal of Social Robotics, 1(1):71–81, 2009.

16. • Hypothesis
 H1: Different proxemics and locomotion behavior will be exhibited with the AR dog
 H2: Different proxemics and locomotion behavior will be exhibited:
• When the dog aware the confederate
• When the confederate aware the dog
 H3: If the dog is aware and reactive to the collision,
• Experience a higher level of co-presence
• Perceive it as a more physical entity
 H4: If the dog is aware and reactive to the collision,
• Score higher in the Animalism category of the Godspeed questionnaire
 H5: Lower levels of affect to the confederate through the affective attraction
• When the dog aware the confederate
• When the confederate aware the dog
Measures and Hypotheses
15

17. • Proxemics and Locomotion Behavior
Results
16
• AR dog influenced proxemics behavior
• No significant difference between groups 𝐶 𝑋,𝐴 and 𝐶 𝑋,𝑈, and groups 𝐶𝐴,𝑋 and 𝐶 𝑈,𝑋
• Effect of social presence

18. • Proxemics and Locomotion Behavior
Results
17
• Slowed down when walking with their AR dog
• No significant difference between groups 𝐶 𝑋,𝐴 and 𝐶 𝑋,𝑈, and groups 𝐶𝐴,𝑋 and 𝐶 𝑈,𝑋

19. • Proxemics and Locomotion Behavior
Results
18
• Significant difference between alone and with dog
• No significant difference between groups 𝐶 𝑋,𝐴 and 𝐶 𝑋,𝑈, and groups 𝐶𝐴,𝑋 and 𝐶 𝑈,𝑋

20. • Proxemics and Locomotion Behavior
Results
19
• Longer path with the dog
• No significant difference between groups 𝐶 𝑋,𝐴 and 𝐶 𝑋,𝑈, and groups 𝐶𝐴,𝑋 and 𝐶 𝑈,𝑋

21. • Proxemics and Locomotion Behavior
Results
20
• Observe more if the dog respond to the collision
• No significant difference between groups 𝐶𝐴,𝑋 and 𝐶 𝑈,𝑋

22. • AR Animal Perception in a Shared Space
Results
21
𝑪 𝑿,𝑨 (respond to collision) shows,
• Higher co-presence score
• Higher associated animalism
• Higher perceived physicality score
• Lower affect was perceived
* Affective Attraction: recognize confederate, 발에 걸려 넘어지지 않았을 경우 confederate을 더 많이 의식함

23. • Qualitative Feedback
 No significant correlation between ownership and how they address AR dog
 Keeping distance not to collide, giving more space to the AR dog
 Check the dog is following well
 Lack of trust if the dog was unaware of confederate
 Looking back to check further interactions
Results
22

24. • Effect of an AR Dog on Proxemics and Locomotion Behavior
 AR dog changed movement and orientation (H1 and H2)
 No attachment sensors → might result in more attentive behavior
• Effect of an AR Dog’s Awareness on Participant Perception
 Reacting to the collision
 AR dog aware the confederate
 Walked over the aware AR dog
 No significant difference whether confederate aware the AR dog (H5-2)
• due to short duration interaction, maybe
 AR dog with awareness and collision response affects,
• the dog’s perceived awareness of the environment
• the perceived awareness of the participant
Discussion
23
→ Feel co-presence (H3)
→ Higher degree of Animalism (H4)
→ Lower affect score to the confederate (H5-1)

25. • Longer duration interaction
• Task oriented AR animal
• AI and IoT will increase interactivity
Future Work
24

26. • Psychological aspects
 Behavior modification
 Positive effects on heartrate or depression
• Good for people who have difficulty raising pets
• Task oriented AR animal
 Dogs for the blind
Criticism
25

27. Thank you😀
Any questions?