For successful collaboration to occur, a workspace must support inter-referential awareness - or the ability for one participant to refer to a set of artifacts in the environment, and for that reference to be correctly interpreted by others. While referring to objects in our everyday environment is a straight-forward task, the non-tangible nature of digital artifacts presents us with new interaction challenges. Augmented reality (AR) is inextricably linked to the physical world, and it is natural to believe that the re-integration of physical artifacts into the workspace makes referencing tasks easier; however, we find that these environments combine the referencing challenges from several computing disciplines, which compound across scenarios. This dissertation presents our studies of this form of awareness in collaborative AR environments. It stems from our research in developing mixed reality environments for molecular modeling, where we explored spatial and multi-modal referencing techniques. To encapsulate the myriad of factors found in collaborative AR, we present a generic, theoretical framework and apply it to analyze this domain. Because referencing is a very human-centric activity, we present the results of an exploratory study which examines the behaviors of participants and how they generate references to physical and virtual content in co-located and remote scenarios; we found that participants refer to content using physical and virtual techniques, and that shared video is highly effective in disambiguating references in remote environments. By implementing user feedback from this study, a follow-up study explores how the environment can passively support referencing, where we discovered the role that virtual referencing plays during collaboration. A third study was conducted in order to better understand the effectiveness of giving and interpreting references using a virtual pointer; the results suggest the need for participants to be parallel with the arrow vector (strengthening the argument for shared viewpoints), as well as the importance of shadows in non-stereoscopic environments. Our contributions include a framework for analyzing the domain of inter-referential awareness, the development of novel referencing techniques, the presentation and analysis of our findings from multiple user studies, and a set of guidelines to help designers support this form of awareness.
| Identifer | oai:union.ndltd.org:GEORGIA/oai:digitalarchive.gsu.edu:cs_diss-1022 |
| Date | 08 August 2007 |
| Creators | Chastine, Jeffrey William |
| Publisher | Digital Archive @ GSU |
| Source Sets | Georgia State University |
| Detected Language | English |
| Type | text |
| Format | application/pdf |
| Source | Computer Science Dissertations |
Page generated in 0.0019 seconds