Evaluation of Interaction Concepts in Virtual Reality Applications

Transcription

1 Maxmlan Ebl, Martn Gaedke. (Hrsg.): INFORMATIK 2017, Lecture Lecture Notes Notes n Informatcs n Informatcs (LNI), (LNI), Gesellschaft Gesellschaft für Informatk, für Informatk, Bonn Bonn Evaluaton of Interacton Concepts n Vrtual Realty Applcatons Hanna Holdered 1 Abstract: Vrtual Realty (VR) s wdely used nowadays and, therefore, t s mportant to provde a hgh usablty of ths technology. In ths paper we evaluate the possbltes of performng devce assessment of prototypes of techncal objects n VR applcatons: We desgned three dfferent nteracton concepts for usng the prototypes. We executed a case study n whch the partcpants had to complete the task of brewng a cup of coffee wth a vrtual coffee machne usng these nteracton concepts. We observed the partcpants behavor durng the completon of the task wth the help of vdeos and questonnares. Two of the nteracton concepts are controller-based, one s a headset-based gaze ponter. The users of the gaze ponter performed best: They completed the task fastest and rated ths concept wth the hghest usablty score. The controller-based concepts were rated lower and the partcpants completed the task slower. The result of our case study can serve as a potental gudelne wth usablty prncples for VR applcatons concernng the desgn of controllers and the dfferent nteracton concepts. Keywords: Vrtual Realty, Usablty, Usablty Evaluaton, Interacton Concepts, Interacton Desgn 1 Introducton Vrtual Realty (VR) s fndng ts way nto more and more areas of the daly lfe. Due to ths development, t s mportant to provde a hgh usablty wth ths technology to enable a satsfyng experence wthn vrtual worlds and ther wde varety of applcatons. The scopes of applcaton are enormous and may provde the possblty of performng devce assessment of prototypes of techncal objects wthout havng the need to have access to the physcal object. Examples are prnters, stereos or, lke n our case, coffee machnes. Therefore, the am of our work s to fnd out whether these devce assessments can be executed properly n VR applcatons and whch nteracton concepts are most sutable for ths purpose. Ths paper sets some basc terms n the begnnng, afterwards we embed t nto the related work n ths area. A descrpton of our nteracton concepts for VR as well as of the case study follow, n the end we evaluate and dscuss the results. 1 Georg-August-Unverstät Göttngen, Insttut für Informatk, Goldschmdtstraße 7, Göttngen, h. holdered@stud.un-goettngen.de c b e do: /n2017_254

2 2512 Hanna Hanna Holdered Holdered 2 Foundatons VR s defned as an artfcal envronment whch s experenced through sensory stmul (as sghts and sounds) provded by a computer and n whch one s actons partally determne what happens n the envronment [Vr17]. Several dfferent approaches such as CAVEs (Cave Automatc Vrtual Envronments), headmounted dsplays or smple devces whch dsplay stereoscopc pctures are used to create VRs [DBJ13]. At the begnnng of our mplementaton n October 2016, the HTC Vve [HTC17], the Oculus Rft and the Playstaton VR were the only VR headsets that used moton controllers and supported postonal trackng [Sh16]. The HTC Vve provded the best 360 degree trackng so we chose a Vve for our case study. It comes wth two controllers whch allow to nteract wth the envronment. The desgn of the controller wth ts buttons and ther names can be seen from dfferent angles n Fgure 1: The trgger wth ts two pressure ponts s on the back of the controller, two grp buttons are postoned on each sde. The large trackpad and the menu and system button are on the front of the controller. In VR applcatons, a vrtual representaton of the controller, whch we call vrtual controller, s shown on the dsplay of the headset. To nteract wth the vrtual envronment, control elements such as the controllers or the headset can trgger certan actons. Those actons are called nteracton tasks and they can only be performed on nteractable objects: After touchng an nteractable object t can be grabbed wth the control element, be attached to and moved together wth t. Those actons correspond and are called touch and grab acton, the latter also ncludes movng the grabbed object. Use actons trgger reactons of the used or a dstant object, e.g. f a lght swtch s used, a lght bulb somewhere else mght turn on. The nteracton technques defne how those nteracton tasks are executed. For example, t s possble to touch an object by movng a control element close to t or by usng raycastng wth a laser beam. For the evaluaton of our case study, we used an adapted System Usablty Scale (SUS). Ths questonnare conssts of ten dfferent statements concernng several aspects of the usablty of the tested system. We slghtly changed some statements to allow us to evaluate the usablty of the grab and use acton. Every statement s rated wth a score from 0 (worst) to 10 (best). To evaluate the whole tested system, the SUS score s calculated from the sum of the scores of the ndvdual statements. The SUS score then ranges from 0 (worst) to 100 (best) ponts. The average SUS score s 69.5 ponts [BKM09]. 3 Related Work There already has been some research n the feld of nteracton technques n mmersve vrtual envronments, especally n late 1990s and early 2000s. Durng that tme, head mounted dsplays were usually combned wth several sensors for trackng a users hand motons [BH99][BH97] and eyes [Ka03], whle controller-based nput devces were not common. The researchers often had to deal wth problems of naccurate trackng hardware, slow system responses, and a hgh latency n general [Ko03]. Ths made t mpossble to

3 proceedngs 2017/8/24 12:20 page 2513 #2513 of Interacton Concepts n Vrtual Applcatons EvaluatonEvaluaton of Interacton Concepts n Vrtual RealtyRealty Applcatons Fg. 1: The desgn of a controller of the HTC Vve wth ts buttons and ther names. focus completely on the nteracton concepts and the results mght have been dstorted. The focus of most case studes les on the nteracton task of grabbng and movng objects, especally dstant ones [BH99][BH97]. Several technques were developed from usablty tests whch showed that ray-castng and extenson of the users arms are not deal for a good usablty [BH99]. The result were technques such as HOMER (Hand-centered Object Manpulaton Extendng Ray-castng) whch uses ray-castng for the selecton and grabbng of the objects. The grabbed object s then attached to a vrtual hand and can be manpulated wth the help of ths hand (hand-centered manpulaton) [BH97]. In [Sc02], several User Interface Concepts of Augmented Realty (AR) Systems are evaluated. Ths topc s closely related to our approach of undertakng case studes for usablty engneerng [Sc02]: In ther approach, Schmdt et al. appled the methods of usng questonnares, ntervews, observng the users nteracton, and thnkng aloud. They descrbed them as deal to get a detaled evaluaton of the system. For our procedure, especally the recordng of the results, we used ths approach of Schmdt et al. as a gudelne. After the early 2000s, there has not been much research n ths area. 4 Interacton Concepts In ths paper, we mplemented three nteracton concepts to evaluate ther usablty. We call them laser controller, arrow controller and gaze ponter. Both the laser and the arrow concept use one controller as control elements whle the gaze ponter s headset-based. VR offers the usage of tooltps on the vrtual controllers. Tooltps are small labels that are connected to a button of the vrtual controller. A text on the label descrbes the buttons functonalty n one or two words. Our two controller-based concepts use tooltps. The nteracton concepts are shown n Fgure 2. Other mplementatons and key assgnments are possble, but based on recent VR applcatons we decded for the followng specfcatons for our case study.

4 42514 Hanna Hanna Holdered Holdered (a) The laser controller wth ts four tooltps, the green laser beam s enabled. (b) The arrow controller wth ts three tooltps, an arrow s attached to ts front. (c) The cross-har of the headset-based gaze ponter n front of the green colored cup. Fg. 2: The appearance of controller-based nteracton concepts laser controller (a) and arrow controller (b) and of the headset-based gaze ponter (c). The tooltps of the controller-based concepts are the black labels that descrbe the functonalty of the related button. 4.1 Laser Controller For the laser controller, a laser beam can be used to pont at and use objects n the VR. The laser beam radates from the top of the vrtual controller and can be actvated by pullng the trgger to the frst pressure pont. Four controller tooltps (see Fgure 2a) descrbe the controller functonalty n ths concept: Ausblenden (Hde) on the menu button: toggle controller tooltps Ponter/Auswahl (Ponter/Selecton) on the trgger: enable laser beam (frst pressure pont) and trgger use actons (second pressure pont) Tasse zurücksetzen (Reset cup) on the trackpad: reset scene to ts orgnal state Grefen (Grab) on grp buttons: grab a grabbable object To touch an nteractable object, the controller has to be moved close to the object untl ts surface colldes wth the surface of the other object. In ths concept, the controller moves through the touched nteractable object. After touchng an object, two thngs happen to ndcate that the object can be grabbed: The touched object changes ts color. The tooltp text of the tooltp Grefen (Grab) changes from whte to red. Both effects can be seen on Fgure 3a wth the arrow concept, whch behaves dentcally n ths case. The text color of the left controller tooltp has become red and the orgnally whte cup has colored grey. Whle the controller touches a grabbable object, t can be grabbed by pressng the grp buttons. After an object has been grabbed, the tooltp text of the grp buttons changes from Grefen (Grab) to Loslassen (Release) to ndcate that the object can be released by pushng the grp buttons agan. The laser beam n ths nteracton concept trggers use actons. When the laser beam s actvated and then hts a usable object, the object changes ts color to green (see Fgure 3b). It changes to ts orgnal color when the

5 Evaluaton Evaluaton of Interacton of Interacton Concepts Concepts n Vrtual n Vrtual Realty Realty Applcatons Applcatons laser beam s moved away. To trgger a use acton when an object s green, the user needs to pull the trgger to the second pressure pont and release t afterwards. 4.2 Arrow Controller In the arrow concept the vrtual controllers have an arrow attached at the front. Ths arrow ndcates that the controller can be drectly used to pont at thngs and to trgger use actons wthout the need to push buttons on the controller. As a consequence, the arrow controller uses only three tooltps (see Fgure 2b): Ausblenden (Hde) on the menu button: toggle controller tooltps Tasse zurücksetzen (Reset cup) on the trackpad: reset the scene Grefen (Grab) on the trgger: grab a grabbable object Touchng objects wth ths nteracton concepts works smlar to the laser concept: the controllers surface has to collde wth an nteractable object whch wll hghlght after t s touched. In contrast to the laser controller, the arrow controller does not move through objects when they are touched but pushes them away. On the one sde ths should emphasze the possblty to use the controller smlar to a hand wthout pressng buttons. On the other sde ths should provde a more realstc feelng because ths s the behavor that would be expected n a real,.e. non-vrtual, world. In ths nteracton concept, the grab acton s trggered by pullng the trgger nstead of pressng the grp buttons whle a grabbable object s beng touched (see Fgure 3a). Another dfference to the laser concept s that the object s attached as long as the trgger s beng pulled. As soon as t s released, the object wll be detached. The adjustment n color of the controller tooltps s the same as for the laser controller. When the tp of the controller s moved closer than a predefned threshold dstance to usable objects, the closest object to the arrow wll be colored green (see Fgure 3c). Ths ndcates whch object wll be used f the controller s moved closer. The use acton tself s trggered when the tp of the arrow hts the green object. Ths behavor should smulate the behavor of a hand n the real world, where the arrow s the ndex fnger whch nteracts wth an object by pushng t. 4.3 Gaze Ponter The gaze ponter can nteract wth the envronment wth a cross-har that s postoned n the feld of vew of the user. To start the equvalent acton of touchng an nteractable object, the cross-har needs to be n front of t (see Fgure 2c). After targetng at an nteractable object t starts to change ts color slowly from ts orgnal color to green as vsble n Fgure 2c. As soon as an object s looked at for a threshold tme of two seconds, the object changes ts color completely and the grab acton s trggered. The grabbed object s then attached to the cross-har and moves along wth t. To release a grabbed object n ths nteracton

6 62516 Hanna Hanna Holdered Holdered concept the user has to break the force that attaches the object to the cross-har. Ths can be acheved by slghtly movng t nto another object. Another varant s that the object has a defned destnaton nto whch t automatcally snaps f a certan dstance to the poston s reached. The use acton works smlar to the grab acton: to trgger a use acton a usable object has to be targeted for a predefned threshold tme of two seconds wth the cross-har. When a usable object lke a button of a coffee machne s beng targeted, t changes ts color to green frst. Durng the next two seconds t slowly becomes red and at the end of the tme span, the use acton s executed. To smplfy the placng of the objects, we mplemented a feature that supports the users n all nteracton concepts: When an object s beng moved nto a predefned area wth a controller-based concept, the color of the tooltp text that says Loslassen ( Release ) changes from whte to red. Ths ndcates that the object should be released there and wll be placed correctly. Wth the gaze ponter, the object automatcally detaches and places tself on the rght spot when t enters the correct area. (a) The grab acton wth the arrow controller: when an nteractable, grabbable object (cup) s touched, the tooltp text color changes from whte to red. (b) The laser controller wth the actvated laser beam, a tooltp (pctogram of an empty cup) of a usable object (coffee machne button) s actvated. (c) The arrow controller n front of a button, the tooltp ( mlder Kaffee ) of the closest usable object (left button of the coffee machne) s enabled. Fg. 3: The grab (a) and use actons (b, c) of the laser (b) and arrow controller (a, c). 5 Case Study To evaluate the nteracton concepts, we performed a case study. The setup of ths case study s descrbed n the followng secton followed by the results. Afterwards, we dscuss the results and the threads to valdty. 5.1 Setup In ths case study, we set up a vrtual room. It conssts of two man objects as shown n Fgure 4: A cup and a coffee machne whch are both located on a bureau. The users task n ths room s to grab the cup and place t at the correct poston on the coffee machne. Afterwards, one of the buttons of the coffee machne whch brews coffee has to be actvated.

7 Evaluaton Evaluaton of Interacton of Interacton Concepts Concepts n Vrtual n Vrtual Realty Realty Applcatons Applcatons Three buttons accomplsh ths task: one for mld coffee, one for strong coffee and one that brews two cups of coffee (framed red n Fgure 4). We decded for a coffee machne because the task of brewng coffee consst of several sub tasks and therefore covers a wde varety of the avalable actons n VR. Furthermore, the real coffee machne s easy to transport for the control group study and t can be assumed that most people have some experences n usng coffee machnes. For carryng out the case study, an nteracton concept s randomly selected by the applcaton for every partcpant. The nteractable objects n ths case study are the cup, whch s grabbable, and the buttons of the coffee machne, whch are usable. To perform ths task, the users have to use the touch, grab, and use acton of the three nteracton concepts. The touch and grab acton are used for movng the cup and the use acton for pushng the buttons of the coffee machne. For documentng our case study, we used several methods such as combnng loggng and recordng vdeos of the screen wth the pcture that the partcpant saw n the VR. We also asked the partcpants to answer a SUS questonnare whch we slghtly adapted to dfferentate between grab and use actons. A data set of the recordngs and results can be found at Fg. 4: The expermental setup: the cup and the coffee machne. The buttons that brew coffee are framed red. 66 partcpants took part n our case study. Every nteracton concept was used 22 tmes. The majorty of our partcpants was male (72.3%), only 27.7% were female. Furthermore, our partcpants were relatvely young: 21.2% were younger than 15 and only 15.2% older than 35 years, n between the dstrbuton was equal. 43.9% of our partcpants clamed that they have no prevous experence wth VR, only 18.2% stated that they are very experenced. We executed a smlar case study wth a control group for the devce assessment of the coffee machne tself. Ths enabled us to dstngush between the usablty problems of the nteracton concepts n VR and the desgn problems of the coffee machne revealed by the devce assessment. In contrast to our expermental group, the control group had to use the real coffee machne whle the task remaned the same: The partcpants n ths group were also asked to brew a cup of coffee. The real coffee machne served as a blueprnt for the vrtual one and had, hence, the same functonalty and desgn. Ths control group study was

8 82518 Hanna Hanna Holdered Holdered less formal. We explaned the task to the partcpants and asked them to brew a cup of coffee whle tellng ther thoughts. The partcpants were observed and ther behavor documented n form of short notes. The control group conssted of seven people from twenty to seventy years old, both male and female. 5.2 Results System Usablty Scale. The overall mean SUS score of our case study s 64.9, whch s slghtly lower than the average score of The gaze ponter reached the hghest score n most statements and n total (68.9), whle the arrow concept was rated worst (59.5). The arrow concept ranks n between wth a total score of Our adapted SUS contaned a statement for each the ntutveness of the grab and use acton. The results of those two statements for every nteracton concept can be seen n Table 1. For the gaze ponter, both statements were rated best of all three nteracton concepts. For the arrow concept, they were rated worst. Laser Concept Arrow Concept Gaze Ponter Queston Mean SD Mean SD Mean SD Grab ntutve Push button ntutve Mean/Total Tab. 1: Results of the System Usablty Scale for the statements concernng the ntutveness of the grab and use acton from all three nteracton concepts. Comparson of the statements between each concept: maxmum (green) and mnmum (red) means hghlghted. Vdeo Analyss. Wth the help of the vdeos we were able to measure the duraton the partcpants needed to complete dfferent sectons of the task. We dvded the task nto sectons: overall duraton from start to completon, from start to grabbng the cup, from grabbng the cup to placng t, from a placed cup to seeng the upper buttons and from that pont of tme to fulfllng the task. The results of ths analyss can be found n Fgure 5, where the tmespans needed to complete the sectons are shown accordng to the nteracton concepts. The x-axs shows the dfferent sectons of nteractons whle the y-axs shows the tme n seconds. The bars represent the nteracton concepts. The gaze ponter users performed fastest n the overall completon of the task as well as n the sectons of grabbng the cup and seeng the upper buttons. In contrast, the users of the controller-based concepts were faster n the placng of the cup. Furthermore, the vdeo analyss allowed us to fgure out general and nteracton concept specfc problems. The general problem that occurred most often s that 14 partcpants dd not see the row of upper buttons that s used to brew coffee, but only the lower left buttons. The problems wth the laser controller are dverse: As far as the grab acton s

9 Evaluaton Evaluaton of Interacton of Interacton Concepts Concepts n Vrtual n Vrtual Realty Realty Applcatons Applcatons Duraton of the Sectons of Interacton accordng to the Interacton Concepts arrow gaze laser tme n seconds start => success start => grabbng grabbng => placng placng => upper buttons upper buttons => success sectons of nteracton Fg. 5: The duraton needed to complete the dfferent sectons of the task. concerned, seven users tred to grab the cup by pressng the trackpad or trgger nstead of the grp buttons. The release of the cup was also dffcult for seven partcpants because they held the grp buttons durng the grab and dd not realze they had to press the buttons agan to release the cup. For the use acton, two other problems occurred wth the laser controller: Ten partcpants tred to trgger the use acton on the coffee machne by movng the controller closer to ts buttons. Another sx partcpants pressed the trackpad to trgger the use acton, whch led to a reset of the cup. For grabbng the cup wth the arrow controller, nne partcpants pressed the trackpad to trgger ths acton. Another eleven partcpants tred to trgger the use acton on the coffee machne by ether pressng the trackpad or pullng the trgger. The only problem that occurred more than twce wth the gaze ponter s that the cup was detached from the cross-har more than fve tmes for fve partcpants because the partcpants broke the force that bound the cup to the cross-har. Control Group. It took the partcpants of our control group between ten to thrty seconds to complete the gven task. As for the qualtatve aspects, they had one man problem: The meanng of the pctogram on the top left button whch shows an empty cup meanng mld coffee dd not seem clear to most of the partcpants. Due to that, they chose to press the next button wth a pctogram of a full coffee cup and therefore brew a strong coffee.

10 Hanna Hanna Holdered Holdered 5.3 Concluson Based on our evaluaton of the SUS score and the vdeo analyss, the gaze ponter performed best and ts users completed the task fastest. The arrow concept ranked lowest and therefore s least usable. The dffcultes that arose wth the control group do not ntersect wth the problems of the expermental group. The only problem of the control group wth the desgn of the pctogram of the top left button dd not occur n the expermental group. Therefore t can be concluded that the usablty problems that occurred wth our applcaton are caused by the nteracton concepts wth the VR and not by the desgn of the coffee machne. Ths also leads to the concluson that our nteracton concepts need to be mproved before VR-asssted devce assessment reveals the same usablty problems as the real coffee machne. Due to the lack of smlar research, we cannot compare our results wth other approaches. The followng secton presents our results for mprovng the nteracton concepts. Grab Acton. The grab acton wth the laser concept receved hgher SUS scores than the arrow concept, but t took the partcpants wth ths concept slghtly longer to complete t. These results can partally be explaned wth the desgn of the rgdness of the cup. As t can be observed n the vdeos, a lot of partcpants pushed the cup because the arrow controller dd not move through t as t dd wth the laser concept. Therefore, the controller should not be rgd and push grabbable objects around. Furthermore, the partcpants expected to trgger the grab acton by pressng the trackpad (arrow concept) or the trackpad and trgger (laser concept). Both buttons seem to be the most promnent buttons of the controllers and therefore are tred frst for nteractons. In consequence, t took the laser concept group longer to fnd the correct button for the grab acton (grp buttons) than the arrow concept group (trgger). Grab actons should therefore be trggered by pressng a promnent button lke the trackpad or trgger. Due to the lack of pressng wrong buttons, no obvous mstake could be observed for the gaze ponter. The hgh score of the SUS statement concernng the ntutveness of the grab acton and the short completon tmes wth the gaze ponter ndcate that less s better: Less controllers and buttons ncrease the usablty of an nteracton. Use Acton. Smlar to the grab acton, the use acton should be trggered by the usage of a promnent button. The lack of a tooltp that told the users of the arrow concept how to actvate a button of the coffee machne mght have led them to try to press any controller button to get the desred result. On the other sde, ten users of the laser concept wanted to actvate the coffee machne by pressng no button at all but by movng the controller close to the coffee machne as t would have worked wth the arrow concept. Therefore, t s up to the developer to decde whether use actons should be trggered by pressng a promnent button or by movng a controller closer to the object that should be used. If the latter s chosen, the controller desgn, e.g. wth an arrow that ponts somewhere, should ndcate ths behavor and dsplay the threshold that wll trgger the acton. The use acton of the gaze ponter s very smlar to the grab acton.

11 Evaluaton Evaluaton of Interacton of Interacton Concepts Concepts n Vrtual n Vrtual Realty Realty Applcatons Placng the Cup and Seeng the Upper Buttons. For placng the cup, the partcpants wth the controller-based nteracton concepts had lttle dffcultes and accomplshed the task much faster than the gaze ponter users. One problem that occurred seven tmes wth the laser concept concerns the release of the cup: Pressng the grp buttons twce, once for grabbng and once for releasng an object, seems to be less ntutve than pressng them just once. A grabbed object should therefore stay attached as long as the grab button s pressed to mmc real world behavor. The problem of the too quckly detachng cup wth the gaze ponter could be solved by ncreasng the force that bnds a grabbed object to the cross-har. Furthermore, the space between the object and ts am should be left free so that the cup cannot ht other objects. In the subtask of seeng the upper buttons of the coffee machne, the arrow concept users were sgnfcantly slower than all other users. Ths could be explaned by the postonng of the coffee machne buttons: After placng the cup, the controller had to be moved upwards to the correct buttons of the coffee machne. The buttons closest to the placed cup where the lower ones. As a consequence, the partcpants tred to nteract wth those buttons. Wthn the laser concept, ths problem dd not occur because the controller was usually held further away from the coffee machne. Ths allowed the partcpants to watch the whole coffee machne and due to ths, they saw the upper buttons as well. The users of the gaze ponter were fastest to see the upper buttons. Ths mght be due to the reason that the gaze ponter s based on lookng at objects closely. 5.4 Threats to Valdty Several crcumstances may have dstorted the results of our case study and due to that some of the results have to be regarded crtcally. The dstrbuton of the nteracton concepts was not perfect even among our partcpants wth ther dfferent precondtons. Intervews and the thnkng aloud technque mght have helped to get a better nsght n the partcpants process of dealng wth the gven task by tellng ther exact thoughts and problems. In addton, our controller layout was chosen wth care but stll there mght be better usable ways to confgure the controller. For example, we dd not mplement a controller-based grab acton usng a laser beam and smlar to ths, other combnatons can be magned. 6 Summary and Outlook In ths paper, we defned three concepts for nteractng wth vrtual prototypes of techncal devces n a VR. Two concepts are controller-based and one gaze-ponter based. The concepts can be used when performng VR-based devce assessment of techncal prototypes. We evaluated the usablty of the nteracton concepts n a case study. In general, the gaze ponter performed best as far as both the completon tme of the task and the SUS score are concerned. Ths can be afflated to ts mnmalst desgn wth no need to learn how to use

12 Hanna Hanna Holdered Holdered the controllers. The controller-based arrow concept performed worst, for the completon tme as well as for the SUS score. The reason can be found n the desgn of the controller and the grab and use acton. Hence, for performng devce assessment of prototypes of techncal objects n VR, the desgn of the nteracton concepts needs to be chosen wth care. An nterestng topc for further research s to evaluate how controller-based nteracton technques can be mproved further, or f controllers can be replaced wth other devces such as the Leap Moton [Lea17]. The Leap Moton s a small devce that can be attached to the front of the headset of the HTC Vve. It tracks the users hand motons and creates a hand model that s dsplayed on the screen of the Vve headsets. 7 Acknowledgments I would lke to offer my specal thanks to Dr. Patrck Harms for gudng me n wrtng my frst scentfc paper. Addtonally, my thanks are extended to the group Software Engneerng for Dstrbuted System wth ts head Prof. Dr. Jens Grabowsk for provdng the hardware. References [BH97] [BH99] Bowman, Doug A.; Hodges, Larry F.: An Evaluaton of Technques for Grabbng and Manpulatng Remote Objects n Immersve Vrtual Envronments. In: Proceedngs of the 1997 Symposum on Interactve 3D Graphcs. I3D 97, ACM, New York, NY, USA, pp. 35 ff., Bowman, Doug A.; Hodges, Larry F.: Formalzng the Desgn, Evaluaton, and Applcaton of Interacton Technques for Immersve Vrtual Envronments. J. Vs. Lang. Comput., 10(1):37 53, [BKM09] Bangor, Aaron; Kortum, Phlp; Mller, James: Determnng What Indvdual SUS Scores Mean: Addng an Adjectve Ratng Scale. J. Usablty Studes, 4(3): , May [DBJ13] Dörner, Ralf; Broll, Wolfgang Grmm, Paul; Jung, Bernhard, eds. Vrtual und Augmented Realty (VR/AR). examen.press. Sprnger Veweg, Berln, [HTC17] HTC Vve. Webste, March 12th, [Ka03] [Ko03] Kaser, Ed; Olwal, Alex; McGee, Davd; Benko, Hrvoje; Corradn, Andrea; L, Xaoguang; Cohen, Phl; Fener, Steven: Mutual Dsambguaton of 3D Multmodal Interacton n Augmented and Vrtual Realty. In: Proceedngs of the 5th Internatonal Conference on Multmodal Interfaces. ICMI 03, ACM, New York, NY, USA, pp , Koutek, Mchal: Scentfc Vsualzaton n Vrtual Realty: Interacton Technques and Applcaton Development. PhD thess, Unversty of Praque, May [Lea17] Leap Moton. Webste, Aprl 30th [Sc02] Schmdt, L.; Oehme, O.; Wedenmaer, S.; Beu, A.; Quaet-Faslem, P.: Usablty Engneerng für Benutzer-Interaktonskonzepte von Augmented-Realty-Systemen. Informatonstechnk und technsche Informatk, 44(1):31 39, 2002.

13 Evaluaton Evaluaton of Interacton of Interacton Concepts Concepts n Vrtual n Vrtual Realty Realty Applcatons [Sh16] Shankln, Wll:, 2016 VR Comparson Gude. Webste, best-vr-headsets-comparson-2016/45984/, June 19th, [Vr17] Vrtual Realty - Defnton. Webste, dctonary/vrtual/%20realty, March 05th, 2017.