Evaluating the effect of positional head-tracking on task performance in 3D modeling user interfaces. (June 2017)

• Record Type: Journal Article
• Authors: Krichenbauer, Max; Yamamoto, Goshiro; Taketomi, Takafumi; Sandor, Christian; Kato, Hirokazu; Feiner, Steven
• Is Part Of: Computers & graphics. Volume 65(2017)
• Page Start: 22
• Page End: 30
• Publication Date: 2017-06
• Subjects: Augmented reality -- Computer aided design -- 3D user interfaces; Computer graphics -- Periodicals; 006.6
• DOI: 10.1016/j.cag.2017.04.002
• Languages: English
• ISSNs: 0097-8493

Abstract

Highlights: Quantitative results of a user study on human factors in 3D user interfaces. Focus on 3D design work (UI & task related to work, participants were artists). The VR device was a modern head-worn display (Oculus Rift DK2). We tested two different input modalities: a 3D input glove and 2D mouse. Found no evidence for effect on human performance from positional head-tracking. Abstract: With the ongoing introduction of wide-FOV VR head-worn displays into the consumer market, the application of VR 3D UIs to professional work environments is attracting increasing attention. One of the most conspicuous concepts is immersive 3D modeling and content creation. In spite of the long research history and multitude of proposed systems, there have been very few analyses of the effect of 3D UIs on productivity in 3D design. In this work, we explore the effect of positional head-tracking and its accompanying parallax depth cues on task performance in 3D object selection and transformation in a setting modeled after 3D design work. Previous studies have come to different conclusions on the importance of positional head-tracking and did not investigate professional 3D modeling tools. In contrast, we performed a user study with design students using professional software on a task that closely emulates their work. Surprisingly, we did not find a significant effect of positional head-tracking on task-completion time, neither when using a traditional 2D mouse nor when using a pinch glove as a 3D input device. Furthermore, we found that the users worked significantly faster with the mouse. We discuss possible explanations and implications for the design of 3D UIs. Graphical abstract: