The Impacts of Animated-Virtual Actors’ Visual Complexity and Simulator Sicknessin Virtual Reality Applications

Iwan Kartiko, Manolya KavakliVISOR Research Group, Dept. of Computing

Macquarie UniversitySydney, Australia

ikartiko, manolya@science.mq.edu.au

Ken ChengDept. of Brain, Behaviour and Evolution

Macquarie UniversitySydney, Australia

ken.cheng@mq.edu.au

Abstract—This article discusses the effects of AnimatedVirtual Actors’ (AVAs) visual complexity on Simulator Sickness(SS) in Virtual Reality (VR) applications. SS is one of the majordisadvantages of VR simulations. Previous research has shownthat visual complexity correlates with SS. Yet complex AVAsare increasingly used along with real-time graphics. MinimisingSS for a VR application is thus beneficial. A series of VRsimulations were created to teach second-year psychologystudents about the navigational capabilities of desert ants withdifferent levels of AVA’s visual complexity: flat, cartoon, or life-like. We predicted that more complex AVAs would induce moreSS. The results contradicted the predictions, with no significantdifferences in SS between groups as a function of the AVAsvisual complexity. Moreover, our methods succeeded in lowoverall levels of SS in all the simulations. Possible explanationsand our future research directions are discussed.

Keywords-virtual actors; virtual reality; visual complexity;simulation sickness; visualisation

I. INTRODUCTION

One of the main goals of in graphics technology andVirtual Reality (VR), is the availability of tools to create low-cost [1], [2], or even free [3], [4] complex Animated VirtualActors (AVAs). This was nearly impossible for low-budgetindustries in the early 1980, but is now readily available.The applications of AVAs for learning in VR extend frommedical education [5] to Risk Assessment Training forcustoms officers [6]. Despite the educational benefits ofVR, the designer must be concerned with the question:”Will it make people sick?” One major disadvantage ofVR applications is that they can cause Cybersickness orSimulator Sickness [7, SS]. The visual complexity of a VRscene is known to induce SS. Therefore, the impact of anAVA’s complexity on SS must be examined to be beneficialfor educational purposes.

In an attempt to minimise the effect of an AVA’s complex-ity on SS, this study employed a recently developed non-interactive VR learning material that was integrated into thecurriculum of second-year psychology students at MacquarieUniversity. This learning material took the students through

Corresponding author: Iwan Kartiko, ikartiko@science.mq.edu.au

a 3D North African saltpan to learn about the navigationalcapabilities of Cataglyphis ants [8]. Depending on the group,they learned with flat, cartoon, or life-like AVAs (Table I).We used standardised the Simulator Sickness Questionnaire[7, SSQ] to measure students level of SS. AVAs are notlimited to humans. The Cataglyphis ants provide complexshapes and tripod-gait information, which is more complexthan a human-shaped AVA. This article defines an AVA’svisual complexity1 as the amount of visual information orelements in depicting an entity.

A. Simulator Sickness (SS)

Motion sickness is widely accepted as a form of sicknessaccording to the Sensory Rearrangement Theory as pro-posed by Reason and Brand [10] and Reason [11]. SensoryRearrangement Theory suggests that symptoms of motionsickness occur when two or more sensory cues disagree, ordo not match with the expected experience. Motion sicknessoften arises from visually-induced self-motion or vection[12], [13]. The symptoms of motion sickness include vertigo,dizziness, stomach discomfort and even vomiting [7]. Theeffect of motion-sickness can be devastating, leading possi-bly to performance interruption and termination of task [14],[15].

In VR applications, it is commonly called Cybersicknessor Simulator Sickness (SS). In VR, the reason for SS liesin the sensory stimuli rather than sensory conflicts (e.g.,[16], [17]) as originally proposed by Sensory RearrangementTheory. So et al. [12] argue that sensory conflicts cannotbe measured directly but sensory stimuli can be measuredreadily. The new trend in the study of SS investigates varioussensory stimuli. The primary stimulus, which is the maintheme of VR applications, is the visual depiction of thescene [12]. Most VR simulations involve the movement ofthe virtual camera, while the user is mostly stationary. Thissense of visual self-motion contradicts vestibular stimuli andtriggers symptoms of SS. Various factors affect the degree

1Visual complexity has various dimensions, such as unintelligibility,disorganisation, amount, heterogeneity, symmetries, colour variety, andthree-dimensionality of elements [9]

2009 Sixth International Conference on Computer Graphics, Imaging and Visualization

978-0-7695-3789-4/09 $25.00 © 2009 IEEE

DOI 10.1109/CGIV.2009.66

147

of visual stimulus in VR that contribute to SS, includingfield of view [18], display size and movement [19], lengthof exposure [20], [21], movement speed [22], and the visualcomplexity of the scene [23], [22], [17], [12].

B. AVA’s Level of Visual Complexity

How does an AVA’s level of complexity affect SS?Commonly in VR applications, the camera is fixed on thestationary AVA (e.g., [5], [6]). In science learning, however,the camera must often follow on AVA to observe a particularbehaviour. Nowadays, creating even more complex AVAsis not difficult with the availability of 3D modelling andpainting tools, along with faster computers and graphic cards[6], [24].

Studies indicate that the complexity of visual stimuli ina scene, arising from scenes with many elements, increasesSS [23], [22], [17], [12]. Kavakli et al. [17] showed thata complex scene with objects, which contain extraneousdetails, such as windows, doors, cracks, and signs of surfaceimperfections, induces more SS than a simple black andwhite scene. In an attempt to quantify visual complexity, Soet al. [12] showed that complex scenes with detailed texturesand additional objects increase SS rating. Similarly, Mourantand Thattacherry [23] and Mourant et al. [22] showed thatadding details such as urban buildings in the scene increasesSS. The increase of SS is expected because higher com-plexity increases the amount of optic flow, creating strongervection [22]. In light of these studies, our initial hypothesisis that the AVA’s visual complexity will also increase overallcomplexity of the scene, which in turn increases vection andproduces stronger SS. The simulation, which was created forPsychology students, contains three different types of AVAs(Table I). We predicted that the SS ratings will follow theorder of AVA complexity.

II. METHODS

A. Participant and Design

The participants were 200 second year psychology stu-dents at Macquarie University (59 males and 141 females)with median age of 20.55 (IQR = 20 - 22). No extra creditwas given as the session was part of the course. A pracgroup of 4 to 25 students participated in the experiment inthe VR lab at the same time. Each prac group as a wholewas randomly assigned to one of 3 conditions differing inthe AVA’s complexity (Table I), 60 participants leant withflat AVAs, 63 participants learnt with cartoon AVAs, and 77participants learnt with life-like AVAs.

B. Material and Apparatus

The Simulator Sickness Questionnaire [7, SSQ] was em-ployed to assess the degree of severity in the simulation.The SSQ was administered before and after the exposureto the VR presentation. The SSQ required the participantsto give a scale from 0 to 3 on each of 16 symptoms listed:

General discomfort, Fatigue, Headache, Eyestrain, Difficultyfocusing, Increased salivation, Sweating, Nausea, Difficultyconcentrating, Fullness of head, Blurred vision, Dizzy (eyesopen), Dizzy (eyes closed), Vertigo, Stomach awareness, andBurping. Questions were summed to obtain three sub-scoresof SS: Nausea, Disorientation and Oculomotor, which can beadded for a total severity score [17]. The VR simulation wasrun on an immersive semi-cylindrical (6m-wide) projectionsystem, which allowed a 160◦ field of view (FOV). TheVR simulation was designed to teach students about thenavigational capabilities of Cataglyphis ants on a featurelesssaltpan [8]. The VR simulation was developed from actualexperiments and observations conducted by Wehner [8],Zollikofer [25] and Wittlinger et al. [26]. It was organizedinto five continuous scenes: 1 - introduction; 2 - overviewof path integration; 3 - skylight compass; 4 - step countingbehaviour; and 5 - summary. The length of the simulationwas 8 minutes and the camera moved slowly from onelocation to another. The virtual actors were visible only inscene 1, scene 2, scene 3, and scene 5. We used Blender[3] for 3D modelling, and animation; Vue [27], GIMP [28]and Inkscape [29] for texturing the virtual world. The VRsimulation ran smoothly at maximum frame-rate. The virtualscene was projected onto the semi-cylindrical screen byusing Vizard [30] at 3027 x 1024 pixels resolution. Thescene was composed of AVAs, plain sand throughout, a clearblue sky, simulated polarisation patterns in the sky, fooditems and three test-channels. Throughout the experiment,the scenario, the scene, screen resolution, screen colour,camera’s movement and the narration were the same. Theonly independent variable is the type of AVA, which wasaltered for each group. None of the students had attended thesimulation before. Since there was no interactivity within theVR simulation, everyone received an equal visual exposure.

Table I summarises the different viewing conditions withthree types of AVA and Figure 1 shows the AVAs on thesemi-cylindrical canvas in our VR Lab. The life-like AVAwas the complex AVA. It was modelled and animated afterthe real Cataglyphis ant. The cartoon AVA was a simplifiedmodel of life-like AVA and composed of fewer polygons,joints and simpler animations. The flat AVA was modelledafter an extreme simplification of a Cataglyphis ant. The flatAVA did not have any 3D shape, only a 2D plane with noanimated joints, and it took less than 2 minutes to create it.

C. Procedure

Upon the arrival in the VR lab, a group of students wasasked to fill in the consent form and randomly assignedto one of three AVA conditions. The students were thenreminded about SS and the procedure to follow if they feelsick in the simulation. SSQ values were collected beforeand after viewing the simulation. Finally, once the studentscompleted the questionnaires, they were thanked for theparticipation.

148

Table IVISUAL COMPLEXITY OF AVAS

Properties

Type Visual Polygons Shape Joints Animation

Simple 1 None None None

Cartoon 1755 Cartoon 18 Human gait

Life-like 2415 Life-like 25 Tripod Gait

III. RESULTS

Initial assessment using the Shapiro-Wilk normality testand the Levene test for unequal variances found violationsfor assumptions needed for parametric statistical tests. Non-parametric tests were thus used in this study. Cliff’s d valueswere calculated to indicate the effect size [31], [32], [33]. Allstatistical analysis methods were performed by using R [34].Table II shows the medians (Mdn) and the interquartile range(IQR) of three sub-dependent measures and the total SSQscore for each group: nausea, disorientation, oculomotor andtotal severity score.

Does the increase of SS follow the order of AVA’svisual complexity? To answer this question, a Kruskal-Wallis nonparametric ANOVA test was performed withAVA’s visual complexity as the factor, and the differencein total SS severity scores (post-exposure score minus pre-exposure score), as the dependent variable. The test didnot reveal any significant differences between the groups,χ2(2, N = 200) = 0.55, p = 0.76. The results did notsupport the proposed prediction and showed no differencesbetween flat, cartoon and life-like AVA in causing SS.

A. Other Results

Additionally, Wilcoxon signed rank test between post andpre total severity scores in each group revealed a significantdecrease of total severity score in all conditions (flat: V =827.5, p = 0.03, d = 0.12; cartoon: V = 887.5, p = 0.02,d = 0.18; and life-like AVA: V = 1449.5, p = 0.00, d =0.15).

Tabl

eII

ME

DIA

NS

(Mdn

)A

ND

INT

ER

QU

AR

TIL

ER

AN

GE

(IQ

R)

OF

SS

QS

CO

RE

SF

OR

TH

RE

EG

RO

UP

S

Pre-

SSQ

Post

-SSQ

Nau

sea

Dis

orie

ntat

ion

Ocu

lom

otor

Tota

lse

veri

tyN

ause

aD

isor

ient

atio

nO

culo

mot

orTo

tal

seve

rity

Gro

upN

Mdn

(IQ

R)

Mdn

(IQ

R)

Mdn

(IQ

R)

Mdn

(IQ

R)

Mdn

(IQ

R)

Mdn

(IQ

R)

Mdn

(IQ

R)

Mdn

(IQ

R)

Flat

6014

.31(

0.00

-28.

62)

13.9

2(0.

00-2

7.84

)18

.95(

7.58

-25.

90)

18.7

0(7.

48-3

8.34

)9.

54(0

.00-

19.0

8)13

.92(

0.00

-27.

84)

15.1

6(7.

58-3

2-22

)13

.09(

3.74

-33.

66)

Car

toon

6319

.08(

9.54

-38.

16)

13.9

2(0.

00-2

7.84

)22

.74(

11.3

7-37

.90)

26.1

8(11

.22-

39.2

7)9.

54(0

.00-

28.6

2)13

.92(

0.00

-27.

84)

15.1

6(7.

58-3

4.11

)18

.70(

3.73

4-37

.46)

Lif

e-lik

e77

19.0

8(0.

00-3

8.16

)13

.92(

0.00

-27.

84)

22.7

4(7.

58-3

7.90

)18

.70(

7.48

-41.

14)

9.54

(0.0

0-28

.62)

13.9

2(0.

00-2

7.84

)15

.16(

7.58

-30.

32)

14.9

6(3.

74-3

3.66

)N

ote.

The

scor

era

nged

from

0to

200.

34fo

rna

usea

scor

e,0

to15

9.18

for

ocul

omot

orsc

ore,

0to

292.

32fo

rdi

sori

enta

tion

scor

e,an

d0

to23

5.62

for

tota

lse

veri

tysc

ore.

149

Figure 1. AVAs as Projected on Canvas

Furthermore, this study discovered gender effects of SS.The Wilcoxon rank sum test was performed, using genderas the factor, and the differences between post and pretotal severity scores as the dependent variable. The analysisrevealed that female participants felt less SS after viewingthe VR simulation than male participants, W = 3262.5,p = 0.02, d = 0.22. There was also a significant genderdifference in the SSQ sub-score: oculomotor, W = 3395,p = 0.04, d = −0.18. However, there was no significantgender difference in disorientation, W = 3558.8, p = 0.09,d = −0.14, or Nausea, W = 3688.5, p = 0.19, d = −0.11.

IV. DISCUSSION

Recent technologies in computer graphics enable artistsand designers to create complex Animated Virtual Actors(AVAs). The availability of the technology enables institu-tions to create AVAs for education and research purposes.

On the negative side, VR is known for inducing SS andone of the main factors of SS is the visual complexityof the scene. Previous studies indicate that the complexityof a scene in VR applications induce SS, which detractsfrom the advantages of the latest technology in creatingcomplex AVAs. Three different types of AVAs were madeavailable for the latest VR learning experience at MacquarieUniversity. The simulation taught students the navigationalcapabilities of Cataglyphis ants. The study investigated theeffects of visual complexity of AVAs in causing SS.

In contrast to the previous studies [23], [12], [22], [17],our findings do not indicate any increase in SS with increas-ing AVA visual complexity. If anything, subjects reportedless SS after viewing the animated scene. We can thinkof several explanations for our results. Firstly, the resultsfrom previous studies in visual complexity and SS are basedon the visual complexity of the virtual environment itself.Kavakli et al. [17] added visual details in the buildings, Soet al. [12] added detailed textures and objects, and Mourantet al. [23], [22] added urban buildings in a driving simulator,and all showed that scene complexity led to higher SS. Thesevirtual environments had complex backgrounds that stoodstill and did not move along with the camera. In our experi-ment, however, the AVAs portrayed in the simulation servedas both foreground and background objects. Secondly, thespeed and the path of the virtual camera were slower anddifferent from the previous studies of SS in VR simulations.We deliberately slowed the camera movements in an attemptto minimise possible SS in the students. This factor deservesto be further tested. Thirdly, the amount of visual complexityof our AVAs might not be sufficient to induce SS. TheAVA’s complexity in this study is small in comparison tothe visual complexity of AVAs in Kavakli et al.s study [17].A fourth possible explanation is that the AVAs do not appearin every scene, and they were sometimes very small becausethe camera was far away.

V. CONCLUSION

In sum, this study assessed the impact of the AVA’s visualcomplexity on simulator sickness. Results showed that thevisual complexity of AVAs did not affect SS, and that theVR simulation on the whole produced very little SS. Wehave made a safe VR presentation for future educationaluse. The results also have important implications for artistsand designers. If the AVA’s complexity does not affect SS,artists and designers can unleash their creativity by usingaffordable and readily available modelling packages, suchas; Quidam [2], Daz3D [1], Makehuman [4] and Blender [3],in creating visually complex AVAs, as long as measures aretaken, as we have done, to minimise overall SS. To confirmthe findings of this study, another simulation is underway totrain fire-fighters in VR.

150

ACKNOWLEDGMENT

This project has been sponsored by the Australian Re-search Council Discovery Grant, DP0558852 (Richards,Kavakli, Dras) titled Risk Management Using Agent BasedSystems. Special thanks to John Porte for his assistance introubleshooting, programming and running the projectionsystem. We are also grateful to Susan Bruck for the fruitfuldiscussion on SS, and Eric Fassbender and Meredith Taylorfor the audio recording session. We are grateful also to NoahBeamer for maintaining Blender Cal3D exporter. We alsothank anonymous reviewers for their comments.

REFERENCES

[1] DAZ3D Team, “Daz 3d,” 2009. [Online]. Available:http://daz3d.com/

[2] Quidam Developers, “Quidam 2,” 2009. [Online]. Available:http://www.n-sided.com/

[3] Blender Developer and User Community, “Blender,” 2009.[Online]. Available: http://www.blender.org

[4] The MakeHuman team, “Makehuman 0.9.1 rc1,” 2009.[Online]. Available: http://www.makehuman.org/

[5] K. Johnsen, A. Raij, A. Stevens, D. S. Lind, and B. Lok, “Thevalidity of a virtual human experience for interpersonal skillseducation,” pp. 1049–1058, 2007.

[6] M. Kavakli, N. Szilas, J. Porte, and I. Kartiko, RiskMan: AMulti-Agent System for Risk Management, ser. ArchitecturalDesign of Multi-Agent Systems. Information Science Ref-erence, 2007, ch. 18.

[7] R. S. Kennedy, N. E. Lane, K. S. Berbaum, and M. G.Lilienthal, “Simulator Sickness Questionnaire: an enhancedmethod for quantifying simulator sickness,” InternationalJournal of Aviation Psychology, vol. 3, no. 3, pp. 203–220,1993.

[8] R. Wehner, “Desert ant navigation: how miniature brainssolve complex tasks,” Journal of Comparative Physiology A:Neoruothology, Sensory, Neural and Behavioral Physiology,vol. 189, pp. 579–588, August 2003. [Online]. Available:http://www.springerlink.com/content/2ewcymk1f6y44heh/

[9] M. N. Roberts, “Complexity and aesthetic preference fordiverse visual stimuli,” Ph.D. dissertation, Departament dePsicologia, Universitat de les Illes Balears, 2007.

[10] J. Reason and J. J. Brand, Motion Sickness. Academic Press,1975, iSBN 0-12-584050-0.

[11] J. T. Reason, “Motion sickness adaptation: a neural mismatchmodel,” Journal of the Royal Society of Medicine, vol. 71, pp.819–829, 1978.

[12] R. H. Y. So, A. Ho, and W. T. Lo, “AMetric to Quantify Virtual Scene Movement for theStudy of Cybersickness: Definition, Implementation,and Verification,” Presence: Teleoper. Virtual Environ.,vol. 10, no. 2, pp. 193–215, 2001. [Online]. Available:http://dx.doi.org/10.1162/105474601750216803

[13] W. Sadowski and K. M. Stanney, Handbook of virtualenvironments: Design, implementation, and applications.Lawrence Erlbaum Associates, Inc., 2002, ch. Presence invirtual environments, pp. 791–806.

[14] C. L. Sinclair, “Motion Sickness - it’s inthe bag!” Touchdown - Royal Australian NavyPublication, December 2003. [Online]. Available:http://www.navy.gov.au/publications/touchdown/dec03/sickness.html

[15] T. Jones, Warefare Officers Career Handbook, T. Jones, Ed.Royal Australian Navy, September 2006. [Online]. Available:http://www.navy.gov.au/publications/warfare/warfare.pdf

[16] N. Bigoin, J. Porte, I. Kartiko, and M. Kavakli, “Effects ofDepth Cues on Simulator Sickness,” ICST IMMERSCOM2007, October, 10-12, Verona, Italy 2007, ACM SIGMMTechnically sponsored by Create-Net and EURASIP, 1-4.

[17] M. Kavakli, I. Kartiko, J. Porte, and N. Bigoin, “Effectsof digital content on motion sickness in immersive virtualenvironments,” 3rd INTERNATIONAL CONFERENCE ONCOMPUTER SCIENCE & INFORMATION SYSTEMS, July23-24, ATHENS, GREECE 2007, 1-12.

[18] R. R. Mourant, N. Ahmad, B. K. Jaeger, and Y. Lin, “Opticflow and geometric field of view in a driving simulatordisplay,” Displays, vol. 28, pp. 145–149, 2007.

[19] H. Shigemasu, T. Morita, N. Matsuzaki, T. Sato, M. Hara-sawa, and K. Aizawa, “Effects of physical display size andamplitude of oscillation on visually induced motion sickness,”in VRST ’06: Proceedings of the ACM symposium on Virtualreality software and technology. New York, NY, USA: ACMPress, 2006, pp. 372–375.

[20] K. M. Stanney and M. Zyda, Handbook of Virtual Environ-ment: Design, Implementation and Applications. LawrenceErlbaum Associates, 2002, ch. Virtual Environments in the21st Century, pp. 1–14.

[21] S. Bruck and P. Watters, The Annual Review Of Cybertherapyand Cybermedicine 2009 - Studies in Health Technologyand Informatics (SHTI) series. IOS Press, Amsterdam,2009, ch. Cybersickness and anxiety during simulatedmotion: Implications for VRET. [Online]. Available:http://www.booksonline.iospress.nl/

[22] R. R. Mourant, P. Rengarajan, D. Cox, Y. Lin, and B. K.Yaeger, “The Effect of Driving Environments on SimulatorSickness,” Proceedings of the 51st Annual Meeting ofthe Human Factors and Ergonomics Society, Baltimore,MD, October 2007, pages 1232-1236. [Online]. Available:http://www1.coe.neu.edu/ mourant/mourant/Publications.html

[23] R. R. Mourant and T. R. Thattacherry, “Simulatorsickness in a virtual environments driving simu-lator,” Proceedings of the IEA 2000/HFES 2000Congress, 1, 534-537, 2000. [Online]. Available:http://www1.coe.neu.edu/ mourant/mourant/Publications.html

[24] M. Kavakli and I. Kartiko, “Avatar Construction for GaitAnalysis,” 3IA’2007 CONFERENCE, THE TENTH INTER-NATIONAL CONFERENCE ON COMPUTER GRAPHICSAND ARTIFICIAL INTELLIGENCE, Athens (GREECE),May 30 - 31 2007.

151

[25] C. P. E. Zollikofer, “Stepping patterns in ants: Influenceof speed and curvature,” Journal of experimentalbiology, vol. 192, pp. 95–106, 1994. [Online]. Available:http://jeb.biologists.org/cgi/content/abstract/192/1/95

[26] M. Wittlinger, R. Wehner, and H. Wolf, “The desert antodometer: a stride integrator that accounts for stride lengthand walking speed,” Journal of experimental Biology, vol.210, pp. 198–207, 2007.

[27] e-on software, “Vue 6 esprit: The art of natural 3d!” August2006. [Online]. Available: http://www.e-onsoftware.com/

[28] Gimp Development Team, “Gnu image manipulationprogram,” 2009. [Online]. Available: http://www.gimp.org

[29] Inkscape Development Team, “Inkscape - open sourcescalable vector graphics editor,” 2009. [Online]. Available:http://www.inkscape.org

[30] Vizard Development Team, “Vizard: Virtual reality toolkit,”2009. [Online]. Available: http://www.worldviz.com

[31] N. Cliff, “Dominance Statistics: Ordinal Analyses to AnswerOrdinal Questions,” Psycological Bulletion, vol. 114, no. 3,pp. 494–509, 1993, ISSN: 0033-2909.

[32] A. B. del Rosal, C. S. Luis, and A. Sanchez-Bruno, “Dominance Statistics: A Simulation Studyon the d Statistic,” Quality and Quantity, vol. 37,no. 3, pp. 303–316, 2003. [Online]. Available:http://dx.doi.org/10.1023/A:1024469826841

[33] J. Romano, J. D. Kromrey, J. Coraggio, and J. Skowronek,“Appropriate statistics for ordinal level data : Should wereally be using t-test and cohen’s d for evaluating groupdifferences on the nsse and other surveys?” Paper presented atthe annual meeting of the Florida Association of InstitutionalResearch, February 1 -3, 2006, Cocoa Beach, Florida, 2006.[Online]. Available: www.florida-air.org/romano06.pdf

[34] R Development Core Team, R: A Language and Environmentfor Statistical Computing, R Foundation for StatisticalComputing, Vienna, Austria, 2008, ISBN 3-900051-07-0.[Online]. Available: http://www.R-project.org

152