Honours Projects
isVR Research Group Honours Project Topics
Requirements Analysis for an Augmented Reality User Interface (Kavakli)
The aim of this project is to explore requirements for an Augmented Reality (AR) Graphical User Interface. An AR interface allows the user to perceive the real world through a head mounted display in which data from a virtual world is augmented over the perceived world. Such interfaces are widely used by the Defense Forces and the Aerospace Industry worldwide. A typical AR interface uses a browser based software package. A VideoServer application (running full screen as an ActiveX Control in Internet Explorer) processes incoming data from a video stream recorded from an externally connected camera. Designing such an interface is challenging, since the virtual world is superimposed on the actual user perception, and interferes with it. Simplifying GUI functionality to a set of minimum core requirements is necessary to increase the task performance. We already developed a training simulation by using a game engine (Unreal Tournament) in our ARC Discovery grant. In this project we will investigate the core requirements for an AR interface integrated into this game engine (Unreal Tournament) and analyse usability engineering aspects, design faults, and user satisfaction in a range of existing AR systems. Based on these results, we will develop a model for an AR interface.
Integration of Motion Tracking into a Generic Virtual Reality Engine (Kavakli)
The VR Engine is a software we have developed integrating a number of software packages with project-specific VR infrastructure. It has a modular system architecture to connect VR hardware and software in a very flexible way, to provide the proper setting for each individual project. The researchers from various departments will have access to the VR infrastructure for their individual projects. The VR Engine integrates the following components: a 3D projection system (to generate stereoscopic view of the virtual world), a video tracking system (DVD cameras and recorders), motion tracking system, game engines, CAD packages (e.g., 3Dmax, Maya and Softimage), and databases with project-specific VR hardware and software. In this project, our purpose is to integrate the input from the software for motion capture (Motion Builder) into our VR Engine. Our magnetic motion capture system works by generating 3 orthogonal electro-magnetic fields form each transmitter. The host graphics station is aware of the timing of the signals. When the receivers pick up the signals, the host knows the distance from the transmitter by the time elapsed and the orientation of the receiver by the changes in the signal caused by tilting of the magnetic fields. Thus, we can utilise the trackers to determine the position as a function of time in 3-D space. Integration of a motion capture suit with a projection cube is a novel idea that will allow the user to freely navigate, immerse, and interact with the objects in a virtual world. Implementation of the VR engine involves the development of an interface between the projection system, VR software, motion capture system, and other VR tools. We have system support personnel responsible for the maintenance and fine-tuning of both VR hardware and software, including installing and integrating various software for specific projects. This may provide guidance to the students who want to work on this integration.
Augmented Prototyping On-Site (Kavakli)
This project involves the use of a vision based augmented reality system. In a vision based augmented reality system, users have access to a combination of Virtual Reality and real world attributes by the superposition of graphical information over the real world. The user continues to operate in the real world but their experience is supplemented by additional information via the virtual interface. The goal of this project is to explore the development of virtual prototyping technologies into practical 3D design/engineering tools through research into the techniques of interaction, visualisation and simulation. The project involves setting up a complex prototyping environment in support of virtual and augmented reality, providing a real time perspective display to the designer/engineer. The project may be extended to include Distributed Virtual Prototyping between geographically separated design teams, to assemble the product drawn using a hand-held force feedback device.
Stereoscopic Heritage Design (Kavakli)
Interactive 3-D graphical interfaces can bring heritage to 'life' enhancing both the educational and the entertainment benefit to the user. Examples of such use of computer generated 3-D graphics to merge the entertainment and education functions are frequently used on television archaeology or science programs when ruined buildings are 'rebuilt' in order to explain the function or construction of the site. This project involves investigating the development of a system for augmented heritage design. The system employs a head mounted display carrying three cameras. Two of the cameras capture a stereoscopic view of the building site, whilst the third is used for viewpoint tracking, in combination with optical markers framing the building site. The image of the building site may either be computer generated or sourced from a remote stereo camera positioned over the user. An immersive projection screen will be used for this project.
Biomechanical Modelling of Human Hand-Wrist Joint (Kavakli)
This project involves the simulation of the human-wrist and hands using physically based modelling techniques to construct a detailed dynamic model of the hand-wrist-joint's behaviour. The project also explores remodelling the bone structure and animating the human wrist and hands. The purpose of modelling the dynamic behaviour of human wrist and hands is to increase the competency and productivity of surgeons, as well as digital artists. Biomechanical studies over a large set of individuals represent an average motion pattern. For motion capture of the hand-wrist joint a pair of data gloves will be used. The application area may be sculpting behaviour, creation of a puppet theatre, or sign language creation for hearing impaired people.
Communication using Speech in Multiplayer Computer Games (Kavakli)
The goal of this project is to develop a collection of speech data (vocabulary, grammar , syntax and semantics) used in a gameplay session, investigating the speech in action between the players. The project involves analysing the video records of game play sessions of a multiplayer game. The data has been collected by using a retrospective protocol analysis technique. Verbal expressions of players will be transcribed from the video tapes and coded by using a coding scheme. This will help us develop a further speech recognition system for a computer game. A Face Tracker to track lip animations will be used for motion recognition and analysis of speech to communicate with the game characters.
Virtual Gallipoli: An Educational War-strategy Simulation (Kavakli)
This project aims at designing a war strategy game based on a navigational map, using a game engine, 3D Modelling packages, and C++ programming. It explores human-computer interaction in a game system using a historical data base. The data about the military strategies, decisions, and their consequences of Anzac war has already been collected by a historian. The project involves the development of a navigational map for making strategic decisions in Anzac War simulation. The navigational map will be used for developing an educational war game. In this project, VR tools such as an immersive projection screen and Spacepad to track the motions of the soldiers are required.
Environmental Simulation for Risk Assessment and Criminal Investigation (Kavakli)
The goal of this project is to produce an accurate real time simulation of the environment to assess the risk of crime. The project involves building a 3D model of Kings Cross for crime investigation and to develop a software to train police officers and architects on risk assessment in a virtual world. A 3D Modelling software will be used to model certain parts of Kings Cross. The environmental simulation will be used for developing a virtual risk assessment kit. New South Wales Police Headquarters already designed a hardcopy version of a Risk Assessment Kit outlining the rules for training. They need a Virtual Risk Assessment Kit to train police officers and architects for risk assessment based on these rules. VR tools such as an immersive cylindrical projector screen, Spacepad and Motion Capture Suit would be used to track motions of the people using the simulation.
Development of a Cyberspace Engine (Kavakli)
This project focuses on building a 3D engine that could have a broad application--namely, cyberspace. In this project, we will explore some possibilities with generalizing 3D game engines into more powerful environments with broader uses. Such a solution would be released under a public license for non-gaming use, so it could fairly rapidly become a standard. This projects involves games and graphics programming.
Analysis of Character Interaction in Computer Games (Hitchens & Kavakli)
Interaction is one of the most enjoyable aspects of playing computer games. Interaction can take many forms within a game, but one of the most important is that with characters others than those directed by the player. Interaction with those characters can include combat, competition, etc. One of the most interesting, but hardest to provide, is conversation. In most existing single-player computer games the possible conversations are pre-programmed. Both what the player can communicate, and what the game controlled characters can respond is pre-set. Players are presented with a list of options, from which they can choose one for the character to "speak". The conversations may be very simple, or extensive, with multi-branching dialog trees. But even in the latter case it is still all pre-determined. Some researchers have approached this problem by attempting to provide full-scale natural language processing systems for us in computer games. Such systems have achieved only limited success, with applications in very restricted settings. They are far from being able to be used in full-scale games. Computer game players do seem to accept the limitations of being able to partake in communication by selecting from a number of options presented by the program. This project will investigate the possibilities of dynamically generating sets of dialog options, rather than having them statically per-determined. The options will be generated from context data, character information and user preferences. The intent is to give the player a more reactive environment, while limiting the programming problem.The work will be based upon the idea of abstracting out the meaning of a communication and using this to first determine the options to be provided to the player and then fleshing them out to full sentences by use of a template database. The project should produce a limited prototype system. A student undertaking this project should have some familiarity with computer games, particularly computer role-playing games, and articificial intelligence (COMP329 or equivalent). Some study in language processing is useful, but not required.