Design a and Assessment o of a Collabor orative 3 3D Interactio ion Techniq hnique f for H Handhe held ld A Augment nted R Realit lity Jerônimo G. Grandi, Henrique G. Debarba, Iago Berndt, Luciana Nedel, Anderson Maciel Graphics, Visualization and Interaction LAB, Federal University of Rio Grande do Sul, Brazil Artanim Foundation, Geneva, Switzerland
2 Augmented Reality Interfaces Mapped Projections Head-mounted See-through Handheld Devices Pixelasm Microsoft Hololens Google ARCore
3 Handheld Augmented Reality Visualization 2D Interaction 3D Interaction Handheld Devices Boring et al., 2010 Grandi et al., 2014 Mossel et al., 2013 Google ARCore Bergé et al. 2014 López et al., 2016 Marzo et al., 2014
4 Handheld Augmented Reality - Characteristics 3D Interaction • Interaction is mapped from the device’s 2D screen to 3D transformations Mossel et al., 2013 • AR allows for position based interaction Marzo et al., 2014
5 Handheld Augmented Reality - Problems 3D Interaction • Still a complex task › Many DoFs › Virtual objects are intangible Mossel et al., 2013 • Exausting if necessary to interact with several objects • Single-user interaction Marzo et al., 2014
6 Collaboration for 3D Spatial Manipulations Divide and Conquer Join and Solve
7 Simultaneous Manipulations of 3D Objects TUI DOF Separation Mean Average Summed Contributions Aguerreche et al. 2010 Pinho et al. 2008 Duval et al. 2006 Grandi et al. 2017
8 Simultaneous Manipulations of 3D Objects Summed Contributions Grandi et al. 2017
9 De Design Asse ssessm ssment and of a Collaborative 3D Interaction Technique for Handheld Augmented Reality
10 The Interface for Collaborative Object Manipulation
11 Des esign Assessment Interface assessment • › Is our technique suitable for Hybrid Interaction: T ouch Gestures and • manipulation tasks in AR Movements environments? Shared space • Collaborative assessement • Simultaneous manipulation • › What strategies emerge when users UI elements are free to make their own task • organization?
12 3D Manipulation Using Handheld Devices for AR Environments
13 Selection
14 Touch Translation
15 Touch Rotation
16 Touch Scale
17 Device Movements
18 Hybrid: Touch + Movements
19 Interface Evaluation Is our technique suitable for manipulation tasks in AR environments?
20 Interface Evaluation - Hypotheses • H1 H1- Hybr brid t tec echniq ique e will be the faste test and will have the lo lowes est er error r rate. • H2 H2- Hybr brid d will have a tim ime a e adv dvantage over the Touch and Movements • H3 H3- Physical effo l effort required by the Dev evic ice m e movem emen ents technique may introduce more re e erro ror r during precise positioning
21 Interface Evaluation Dependent Evaluation Details Independent Variables Variables Touch Gestures Time Device Movements Accuracy Hybrid
22 Interface Evaluation Dependent Evaluation Details Independent Variables Variables Task Subjects Setup & Protocol x 20 volunteers • Touch Gestures Time 14 male • 6 female • repeated measures • Device Movements Accuracy within-subject • 8 trials per condition • Participants had to • confirm to finish the Combined docking
23 Results - Performance ** *** 1.15cm and 8 ◦ Participants had to confirm to finish the docking The picture can't be displayed. H2- Hybr brid will have a time a e advan antag age e over the Touch and Movements
24 Results - Error reduction by time for position and rotation The picture can't be displayed. H1- Combi bined t d techniqu que will be the fastest st and will have the lowest e err rror ra rate.
25 Results - Workload ** ** ** ** H3- Phys ysical e effort required by the Devi vice m move vements technique may introduce more e error during precise positioning
26 Collaborative 3D Manipulation
27 Interface Elements
28 Simultaneous Manipulations of 3D Objects Summed Contributions Grandi et al. 2017
29 Collaborative Evaluation What strategies emerge when users are free to make their own task organization? How the performance changes accordingly to the strategy adopted?
30 Interface Evaluation - Hypotheses • H4 H4- Two strategies will emerge: Indep epen enden ent I Inter eractio ion and Shared I Inter eractio ion • H5 H5- Users will tend to reo eorganiz nize e them emselv lves d dep epend endin ing o on the t task`s difficult ulty • H6 H6- Sha hared int interaction approach will have better time performance compared with Independent Interaction.
31 User Study – Main Hypothesis Dependent Independent Variables Variables Low Occlusion Total Time Moderate Occlusion User Maniputation Time High Occlusion
32 Assessment Conditions User 1 View User 2 View
33 User Study – Main Hyphotesis Dependent Evaluation Details Independent Variables Variables Task Protocol Subjects repeated measures • within-subject • 8 trials per condition • x15 Low Occlusion Total Time 1.15cm and 8 • volunteers degrees finishes the • 17 male dock • 13 female • Moderate Occlusion User Maniputation Hybrid Manipulation Time High Occlusion
34 Results – Pair Performance
35 Results – Pair Strategies H4 H4- Two strateg egies w will appea ear: I : Indep epend endent ent I Inter eraction n The picture can't be displayed. and S Shared ed Inter eraction. n. H5- Users will tend to organize t H5 themselves d depending on n the he t task`s d difficulty. Pairs s stick w with t thei eir s strategy i indep epen enden ent o on the e occlusion H6 H6- Shared ed i inter eracti tion n approach will have better er t time e perform rmance compared with Independent Interaction. Groups
36 Results Summary • Handheld devices are po powerfu ful tools for 3D collaborative tasks in AR
37 Results Summary • Handheld devices are po powerfu ful tools for 3D collaborative tasks in AR • The Hybrid t d techniqu ique e achieves the best performance
38 Results Summary • Handheld devices are po powerfu ful tools for 3D collaborative tasks in AR • The Hybrid t d techniqu ique e achieves the best performance • Two s strateg egies ies were adopted Groups
39 Results Summary • Handheld devices are po powerfu ful tools for 3D collaborative tasks in AR • The Hybrid t d techniqu ique e achieves the best performance • Two s strateg egies ies were adopted • Both strategies performed in similar Groups time
40 Applications Serious Games Assembly tasks Data Exploration Ford VisNEST Medtech Boston ESI Group CAVE2™ CNN T ech
41 Future Work • Add more subjects in the scene • Control the labor division • Remote locations • Explore other immersion scenarios
Design and Evaluation of a Handheld-based 3D User Interface for Collaborative Object Manipulation Jerônimo G. Grandi, Henrique G. Debarba, Iago Berndt, Luciana Nedel, Anderson Maciel jggrandi@inf.ufrgs.br Handheld devices are power erful ul tools for 3D collaborative tasks in AR The Hybri rid d technique achieved the bes est p per erformanc nce Indep epen enden ent I Inter eraction and Shared I Inter eractio ion were the strategies adopted by pairs Performan ance ce during pair work is not r rel elated ed to the strat ategy adopted Graphics, Visualization and Interaction LAB, UFRGS, Brazil Artanim Foundation, Geneva, Switzerland
Recommend
More recommend
