- Title
- PREVENT–A pipeline approach to prototype realistic virtual environments via the reuse of expert domain knowledge
- Creator
- Xi, Mingze
- Relation
- University of Newcastle Research Higher Degree Thesis
- Resource Type
- thesis
- Date
- 2017
- Description
- Research Doctorate - Doctor of Philosophy (PhD)
- Description
- Building realistic virtual environment-based training systems has been a long-term research topic. One issue is the difficulties of embedding expert domain knowledge into the virtual environments. For example, many virtual fire evacuation systems have limited fire science knowledge, resulting in unrealistic evacuation behaviours of computer-controlled virtual humans and inaccurate fire modelling. This research project has developed and validated a pipeline approach, PREVENT (a Pipeline for pRototyping Evacuation training Virtual ENvironmenT), which reuses domain simulators and game engines to create realistic virtual environments. PREVENT has been shown to be consistent, accurate, and scalable in a series of case studies. In addition to the reuse of domain knowledge, an interaction framework was designed and demonstrated to effectively enable dynamic interaction in virtual environments created via PREVENT. To evaluate the behavioural realism of the virtual humans, a new experimental protocol was designed based on the conventional Turing Test. A user study demonstrated that the virtual humans generated via PREVENT behaved as realistically as real human participants in an example fire evacuation drill environment.
- Subject
- virtual environment; non-player characters; software pipeline; virtual fire drill; turing test; intelligent behaviour; games
- Identifier
- http://hdl.handle.net/1959.13/1353325
- Identifier
- uon:31084
- Rights
- Copyright 2017 Mingze Xi
- Language
- eng
- Full Text
- Hits: 1013
- Visitors: 1959
- Downloads: 853
Thumbnail | File | Description | Size | Format | |||
---|---|---|---|---|---|---|---|
View Details Download | ATTACHMENT01 | Thesis | 14 MB | Adobe Acrobat PDF | View Details Download | ||
View Details Download | ATTACHMENT02 | Abstract | 131 KB | Adobe Acrobat PDF | View Details Download |