CITEC Central Lab Facilities
CITEC
British Flag   German Flag

3D User Interfaces Conference 2010 - The Grand Prize Contest

The 3D User Interfaces Conference 2010 had a call for a 3D UI Contest: The 3D UI Grand Prize. The goal was to find innovative solutions to the classic 3D UI problems of

They provided a supermarket scenario file coded in VRML and a task: the user should collect three distinct objects from the shelves (navigation and selection) and place them on a table in a certain orientation (manipulation).

The performance of the solution should be measured in terms of time for task-completion and average FPS.

They also offered two contests categories: on-site and video-only.

Contribution of the Interaction in Virtual Reality project group @ AG WBS/UniBi

Our group decided to participate in the video-only category with our student group of 3D UI freshmen in the fall/winter semester of 2009/2010.

Our Vision

We wanted to implement well-known techniques for navigation, selection and manipulation using up-to date sensor devices and see, how they compete in this scenario. Doing so, we would also provide a baseline for other contestants with innovative new techniques.

The Challenge

The students were 3D UI freshmen, most of them without previous experiences in 3D programming let alone 3D User Interface design.

Our Advantage

The students were highly motivated and enthusiastic to use our 3-sided CAVE and experiment with our interaction devices.

The Results

Five combinations of techniques for selection/manipulation and navigation have been tested:

Time-on-Task/Experts

The contest required that three objects should be picked-up in the supermarket and placed on a table in distinct orientation and position. The following table summarizes the mean time on task for expert users:

Condition Number of Participants Time on Task
meanmin max
Ray-Casting/PathDrawing 11 7:07 min 3:08 min 13:16 min
Ray-Casting/Walking in place 11 5:37 min 2:32 min 9:03 min
Ray-Casting/World in Miniature 7 6:16 min 3:00 min 9:58 min
Image Plane/Lean-based Velocity
(only one object)
7 2:10 min 1:00 min 4:57 min
Image Plane/Grabbing in the Air
(only one object)
6 2:24 min 0:47 min 4:00 min

 

Questionnaire/Novices

We could also tell an interesting story about testing these approaches on really naive people. We started the hard way by testing 20 students from other faculties (humanities, natural sciences, no technicians, no computer game players). However, this kind of failed, as it took them half an hour or more to solve the task (one object) with even the simplest solution. So in terms of performance/timing we finally decided to run the study again with experts, that is, people with prior experiences in VR.

We did, however, run a questionnaire afterwards to test

of the individual methods (rated on a scale from 1 to 5) and these results are presented in the video and the description of the individual techniques, besides the performance of the experts.

Video

The following video, which has been submitted to the 3D User Interface Grand Prize, summarizes those five combinations, developed by our student team. We hope you enjoy their final presentation as much as we enjoyed working with them:

The full video can also be downloaded here: