Full-Body Interaction-based Learning Support to Enhance

Immersion in Zoos

Evaluating an Electrodermal Activity Response Support System

Mikihiro Tokuoka

, Hiroshi Mizoguchi

, Ryohei Egusa

2,3

, Shigenori Inagaki

and Fusako Kusunoki

Department of Mechanical Engineering, Tokyo University of Science, 2641 Yamazaki, Noda-shi, Chiba-ken, Japan

JSPS Research Fellow, Tokyo, Japan

Graduate School of Human Development and Environment, Kobe University, Hyogo, Japan

Department of Computing, Tama Art University, Tokyo, Japan

Keywords: Body Movement, EDA, Kinect Sensor, Learning Support, Virtual, Zoo.

Abstract: We are developing a learning support system for zoos that enables children to learn through body

movements. For children, the zoo is an important place for science education outside of school. However,

learning methods in zoos are mainly limited to observing exhibits and explanations. In addition, when

children want to observe animals that are hiding, they can neither touch nor observe them. Therefore, in this

study, we develop a body-experience-based learning support system that can be applied in zoos. The system

measures the learners’ body movements using a sensor, and provides a sense of immersion in the

environment where animals live by adapting the surroundings based on these movements. In the first stage

of this project, we have developed a prototype that allows learners to jump with animals, and enables

children to efficiently learn animals’ physical characteristics. In this paper, we describe the results of our

experimental evaluation of by measuring physiological responses.

1 INTRODUCTION

Hundreds of millions of visitors worldwide, most of

whom are children, experience zoos each year

(Wagoner, 2010). For these children, the zoo is a

very important place for science learning outside of

the classroom (National Research Council, 2009);

observing and touching animals motivates children

to learn more about the animals and their ecology in

detail (Braund, 2006). However, learning methods in

zoos are mainly limited to observing exhibits and

listening to explanations. Furthermore, children can

neither touch nor observe animals that are hiding,

making it impossible to make contact with the

animals; it is very difficult for children to imagine

the ecology of animals by themselves, which

reduces their motivation to study. To solve these

problems, conventional research has proposed

learning support, including providing video content

and using tablets (Zarzuela, 2013; Webber, 2015;

Tanaka, 2017). However, these studies lack active

mass and children’s experiences; they are far

different from the actual experience of making

contact with animals, presenting a fundamental

problem. Solving these problems requires a system

that enables children to observe animals and have an

authentic experience.

Children think hard while playing (Dau, 1999;

Levin, 1996). When children use gestures and

movements, the learning environment becomes more

natural (Grandhi, 2011; Nielsen, 2004; Villaroman,

2011) and children can retain more of the knowledge

being taught (Edge, 2013; Antle, 2009). Therefore,

we focus on playing while moving the body; by

using body movements, we can provide real

experiences and impart knowledge.

We are developing a learning support system for

zoos that enables children to learn through body

movements. In this system, body movement

information is acquired by a sensor, and the content

operates based on this information. We attempt to

provide realistic experiences of animals that are

more difficult to contact directly. Furthermore, by

reflecting the learners’ body movements in virtual

space as an observation behavior, the learner feels

immersed, and their experience approaches

336

Tokuoka, M., Mizoguchi, H., Egusa, R., Inagaki, S. and Kusunoki, F.

Full-Body Interaction-based Learning Support to Enhance Immersion in Zoos.

DOI: 10.5220/0006768703360341

In Proceedings of the 10th International Conference on Computer Supported Education (CSEDU 2018), pages 336-341

ISBN: 978-989-758-291-2

authenticity; we therefore expect that they can learn

sufficiently about the ecology of animals.

In this paper, we describe the results of

evaluating the usefulness of this system by

developing and evaluating prototypes as a first step

in realizing a system to support learning about

animals in the zoo. We quantitatively evaluate

interest by measuring electrodermal activity (EDA),

which is physiological response that indicates

whether children are interested in the animals and

motivated by the system.

2 LEARNING SUPPORT IN ZOOS

2.1 System

We are developing a learning support system for

zoos that enables children to learn through body

movements. We use the example of the wildcat,

which is a felid belonging to the class Mammalia.

Wildcats are shy and do not commonly interact with

people. However, wildcats can jump high. In zoos,

wildcats hide, because they are not good at

connecting with people; therefore, they are difficult

for children to observe. Our proposed system

supports learning about wildcats by using various

sensors to measure a learner’s position, attitude and

movement, and operating a learning system based on

these data. Figure 1 shows a schematic of the

system; in the space displayed onscreen, the object

to be learned changes in conjunction with body

movement. The system provides learners with the

feeling of being in the environment where the

animals live. In addition, real observations of

behavior—approaching animals, jumping, etc.—are

incorporated to give learners a more realistic

experience than just watching exhibits or videos. It

is thus possible to increase their motivation to learn

and the knowledge obtained. As the animals move

according to the children’s body movements, the

learners identify the animals as realistic, rather than

just virtual creatures.

2.2 System Configuration

To implement our zoo learning support system, we

are developing a system that lets children learn

about ecology of the wildcat while projecting them

to an onscreen grassland to jump with a wildcat.

In this system, when a student stands in front of

the screen, the region containing a person is

extracted, and the learner’s body is projected

Figure 1: System concept.

Figure 2: System setup.

with a wildcat by operating the system using body

movements.

The height of the learner's jump is measured by a

sensor, and the learner also jumps along with a

wildcat on the screen; in this way, learners acquire

knowledge of the wildcat’s jumping ability.

Figure 2 shows that this system consists of a

Microsoft Kinect sensor, control PC, and projector.

The Kinect is a range-image sensor originally

developed as a home videogame device. Although it

is inexpensive, the sensor can record sophisticated

user location measurements. Additionally, this

sensor can recognize humans and the human

skeleton using a library in its Windows software

development kit. The Kinect can measure the

location of human body parts such as hands and

legs, and can identify the user’s pose with this

function and its location information. The Kinect

can measure the three-dimensional coordinates of

the skeletons of 25 human beings. To ascertain the

height of a jump, we measure the coordinates of the

learner’s ankle. Next, we extract the image region

containing a human. In Kinect's human body

database, various human body posture patterns

Full-Body Interaction-based Learning Support to Enhance Immersion in Zoos

337

are machine learned and the parts of the body are

identified using this database.

The functions and flow of the system are shown

below:

(a) Operates using the learner’s body movement

(b) Projects a person onto the screen

(d) Teaches animal characteristics.

Function (a) allows the system to operate as directed

by the user’s hand movements. A learner can click a

button by pushing their hand toward the Kinect

sensor; the learner starts the game with this action.

Clicking the start button brings up a screen for

jumping with a wildcat. The learner can start the

game with their body as an interface. Function (b)

displays the user on the screen by recognizing the

sensor region the learner is in. Learners can immerse

themselves in the world of a wildcat on the screen.

Function (c) allows the user to jump with a wildcat

on the screen. The wildcat on the screen mimics the

height at which the learner jumps. Learners would

be unable to interact in this way with a real wildcat

in the zoo, but our jumping system enables learners

to develop familiarity with the wildcat. Function (d)

gives learners the experience that wildcats are good

at jumping. The system digitizes the actual jump

heights of the learner and the wildcat. When the

learner jumps with all their power, the wildcat is

shown at its impressive jump height of around 2 m.

Therefore, the learner can experience the fact that

wildcats are good at jumping. Figure 3 shows the

system flow and its state while performing Functions

(a), (b), (c) and (d).

3 EVALUATION

3.1 Objective

In this study, we used the method of immersing the

learner in a virtual environment and jumping with a

wildcat to learn about the ecology of wildcats. We

hypothesized that learners’ interests would be

piqued by this environment and activity. However, it

is possible that children are simply interested in

jumping. Therefore, we examined whether children

are more interested when jumping with a wildcat or

jumping alone. We evaluated interested

quantitatively using electrodermal activity (EDA)

which is a physiological response. We thus aimed to

evaluate the usefulness of our system.

Figure 3: System flow.

CSEDU 2018 - 10th International Conference on Computer Supported Education

338

3.2 Evaluation by EDA

3.2.1 EDA

We used EDA to confirm the presence or absence of

children’s interest. EDA is the change in the

electrical properties of the skin’s surface due to

sweating caused by excitation or tension. The

method of measuring apparent resistance by passing

electricity through the skin is referred to as the

energization method. The sustained activity

measured thereby is referred to as skin conductance

level (SCL). Transient activity is referred to as the

skin conductance response (SCR).

The brain's limbic system is comprised of the

cingulate gyrus, hippocampus, and amygdala.

Recent studies have demonstrated that the limbic

system is activated when a person is interested. In

addition, this system is known to have a close

relationship with emotion (Boucsein, 2012). When

people are visually interested in a subject, they

experience a pleasant feeling and an increase in

arousal, which changes the electrical activity of the

skin and produces an SCR (Yoshida, 2014). Figure 4

shows that if interest is generated, the SCL and SCR

reaction values can be measured from variations

over relatively long and short periods, respectively.

When interest occurs, a reaction appears. By

contrast, no reaction appears when there is no

interest. Therefore, we judge the presence or absence

of interest in the proposed system using EDA.

3.2.2 Method

Twelve fifth- and sixth-grade students (7 boys and 5

girls) from elementary schools attached to the

national university corporation participated in an

experiment at the Hyogo Prefectural Museum of

Natural History on October 27–29, 2017.

We attached an electrode that measures EDA to

a fingertip on each subject’s right hand, as shown in

Figure 5. The participants each experienced the

system individually.

First, children experienced the proposed system

for jumping with a wildcat. Next, they simply

jumped. We measured fluctuations in EDA with

each experience.

3.3 Result

First, we examine representative experimental

results from one user. Figure 6 shows a graph of the

change in EDA when experiencing the proposed

system, whereas Figure 7 shows a graph of the

change in EDA when simply jumping in front of the

screen. The horizontal axis of the graph is time

Figure 4: EDA.

Figure 5: Experimental environment.

Figure 6: EDA in the proposed system experience.

variation and the vertical axis is the degree of

change. From Figures 6 and 7, we see that interest

occurs when users experience the proposed system.

We evaluated our results by calculating the

number of occurrences of interest in each result

compared to the number of trials. When jumping

with a wildcat, interest occurred 29 times. However,

when simply jumping, interest was demonstrated 6

times. This result quantitatively confirmed that

learners are not interested in jumping, but that they

are interested in jumping with a wildcat. Thus, it was

possible to quantitatively evaluate the number of

occurrences of interest.

Next, we considered the experimental results of

all the subjects, which are listed in Table 1.

Statistical tests were conducted on the EDA results

Full-Body Interaction-based Learning Support to Enhance Immersion in Zoos

339

in the two scenarios, and a t-test analyzing the

difference between the average EDAs produced a

result of p < 0.01, which corresponds to a significant

difference.

With this result, we have demonstrated

quantitatively that children are interested in the

wildcat and motivated to learning about the wildcat

after experiencing the proposed system. Figure 8

shows children experiencing the proposed system.

Figure 7: EDA when simply jumping.

Table 1: Results of all subjects.

number of EDAs

Subject

Proposed system

Jumping only

Figure 8: Learner in the system experience.

4 CONCLUSIONS

In this paper, as a first step in implementing a

learning support system for children visiting zoos,

we proposed a system that allows children to jump

together with a wildcat using a Kinect sensor. We

also described the quantitative evaluation results of

measuring EDA, which is physiological response, as

a metric of whether children are motivated by and

interested in virtual contact with animals.

From our experiment, it became quantitatively

clear that learners can be interested in animals and

motivated by the proposed system experience, which

enables them to jump with virtual animals on a

screen following their body movement. This result

demonstrated that the proposed system effectively

provides a platform for observation of and learning

about the wildcat, an animal that is difficult to

observe directly in a zoo.

In the future, it will be necessary to evaluate

knowledge acquisition and conduct experiments at

real zoos. First, children observe the system after

observing a rooster at the zoo. Then, we will

evaluate the acquisition interests and knowledge,

which with further the system’s development. This

experiment is already scheduled.

ACKNOWLEDGEMENTS

This work was supported in part by Grants-in-Aid

for Scientific Research (A). The evaluation was

supported by the Museum of Nature and Human

Activities, Hyogo, Japan.

REFERENCES

Wagoner, B., Jensen E., 2010. "Science learning at the

zoo: Evaluating children’s developing understanding

of animals and their habitats." Psychology & Society,

65–76.

National Research Council, 2009. Learning science in

informal environments: People, places, and pursuits.

National Academies Press.

Braund, M., Reiss, M., 2006, "Towards a more authentic

science curriculum: The contribution of out‐of‐school

learning." International Journal of Science Education,

1373–1388.

Zarzuela, M. M., et al., 2013. "Mobile serious game using

augmented reality for supporting children's learning

about animals." Procedia Computer Science 25, 375–

381.

Webber, S., 2015. "Design and evaluation of interactive

technology for human–animal encounters at the zoo."

CSEDU 2018 - 10th International Conference on Computer Supported Education

340

Proceedings of the 12th International Conference on

Advances in Computer Entertainment Technology,

ACM.

Tanaka, Y., Egusa, R., Dobashi, Y., Kusunoki, F.,

Yamaguchi, E., Inagaki, S., Nogami, T., 2017.

“Preliminary evaluation of a system for helping

children observe the anatomies and behaviors of

animals in a zoo.” Proceedings of the 9th International

Conference on Computer Supported Education 2,

305–310.

Dau, E., Jones, E., 1999. Child's Play: Revisiting Play in

Early Childhood Settings. Brookes Publishing, Maple

Press.

Levin, D., 1996. “Endangered play, endangered

development: A constructivist view of the role of play

in development and learning.” In A. Phillips (Ed.),

Topics in early childhood education 2: Playing for

keeps. St. Paul, MI: Inter-Institutional Early Childhood

Consortium, Redleaf Press

Grandhi, S. A., Joue, G., Mittelberg, I., 2011.

Understanding naturalness and intuitiveness in gesture

production: insights for touchless gestural interfaces.

Proceedings of CHI, 821–824.

Nielsen M., Störring, M., Moeslund, T., Granum E., 2004.

“A procedure for developing intuitive and ergonomic

gesture interfaces for HCI.” Gesture-Based

Communication in HCI, 105–106.

Villaroman, N., Rowe, D., Swan, B., 2011. “Teaching

natural user interaction using OpenNI and the

Microsoft Kinect Sensor.” Proceedings of SIGITE,

227–232.

Edge, D., Cheng, K. Y., Whitney, M., 2013. “SpatialEase:

learning language through body motion.” Proceedings

of CHI, 469–472.

Antle, A., Kynigos, C., Lyons, L., Marshall, P., Moher, T.,

Roussou, M., 2009. “Manifesting embodiment:

Designers’ variations on a theme.” Proceedings of

CSCL, 15–17.

R. Yoshida, T. Nakayama, T. Ogotsu, H. Takemura, H.

Mizoguchi, E. Yamaguchi, S. Inagaki, Y. Takeda, M.

Namatame, M. Sugimoto, F. Kusunoki., 2014.

“Feasibility study on estimating visual attention using

electrodermal activity,” Proceedings of the 8

International Conference on Sensing Technology.

Boucsein, W., 2012. Electrodermal activity. Springer

Science & Business Media.

Full-Body Interaction-based Learning Support to Enhance Immersion in Zoos

341