School of Computing & Mathematics

DP130100364: A probabilistic framework for nonlinear dimensionality reduction algorithms
Description:The Twin Measures Framework is a novel platform for analysing existing dimensionality reduction methods and the invention of new ones.  This research will radically improve image analysis, with beneficial applications from pharmaceutical drug design through to border protection.

Prof Junbin Gao and Prof Xia Hong (Univ of Reading, UK)

DP130103670: Multiview video coding using cuboid data compression
Description:  This project investigates novel approaches to multiview video coding that use new data compression techniques and explicit occlusion handling.  These new approaches complement the state-of-the-art, improving interactivity with instantaneous view change and VCR functionality, reducing encoding complexity and increasing compression efficiency.

Prof Manzur Murshed (Monash) and Dr. Manoranjan Paul

Investigating which strategies are most effective in overcoming the ethical problems facing information and communications technology professionals
Description:  This project will investigate what sorts of ethical problems information and communications technology (ICT) professionals commonly face at work and which strategies are most effective in solving them.  An output of this project will be an interactive multimedia-rich website that will enable ICT professionals to deliberate on an ethical problem and come up with the best option for solving it.

Crisis management simulation: developing a methodology for transforming communication response
Description:  This project merges cutting-edge digital games technology with applied drama techniques to produce a crisis management game to simulate conflict and crisis scenarios.  Working closely with the Australian Defence Force to better understand organisational communication under extreme pressure, this new approach will build teamwork and break down barriers to effective crisis management.

WAREIGS: Wavelet-Network-Based Augmented-Reality-Enhanced Image-Guided Surgery
Description:  This research will address issues with the constraint of surgical openings where surgeons cannot see beyond the exposed surfaces and these limitations are accentuated by the even greater restrictions of minimally invasive surgery.  Limited visibility through "keyholes" during endoscopic procedures and through small incisions with ever-diminishing sizes increases the need for intra-operative image guidance.  the project aims to develop a system of augmented reality enhanced image guided surgical surgery, in which the images from cameras are aligned with patient's physical position at the time of surgery.  Such a "see-thorugh" capability makes surgeries safer and more accurate.  two objectives will be achieved: (1) Theoretical research on image registration based on wavelet networks.  (2) Development of an sugmented reality enhanced image-guided surgery with see-through capability.

An evaluation of nurse practitioners – aged care models of care program in Australia
Description:  The purpose of the Nurse Practitioner Aged Care Models Program is to develop effective, viable and sustainable models of service, grow Nurse Practitioner numbers, and improve access to primary aged care service.  Amongst the aims to the national evaluation of this Program is to assess the impact of new models in relation to cost-effectiveness, quality of care and health outcomes.

Investigation of Imaging Techniques to Determine Muck Pile Ore Fragment Size In-Situ
Description:  We investigate real-time techniques for detection of larger size ore fragment in muck pile under the production environment.

The threats from Data Mining on the privacy of a Facebook User
Description:  In this project we discover various possible ways to breach the privacy of an individual Facebook user by a malicious data miner using sophisticated data mining algorithms.  The main goal of the project is to find out the threats on privacy and increase public awarements.  We also argue that it is a responsibility of the data miners to provide a technical solution to the problem.  We also propose data mining solutions in order to protect the individual privacy of a Facebooker user from the threats of a malicious data miner.

Efficient design for Generalised Linear Models
Description:  Generalised Linear Models (GLMs) provide an extention of the traditional regression and Analysis of Variance models based on the Normal distribution to other distributions.  Designing efficient experiments for these traditional models is comparatively stratightforward because the optimal design does not depend upon the values of the unknown parameters of the linear model.  However, in GLMs, the optimal design does depend upon the unknown parameters, which makes the problem of finding efficient designs more difficult.

A Concurrent Hybrid Approach to Secure Computer Systems
Description:  This project investigated concurrent application of AI and visualisation techniques for network security.

Twin Measure Framework for Dimensionality reduction
Description:  This project aimed at developing a family of efficient nonlinear dimensionality reduction algorithms for data visualisation.

3D Video Coding
Description:  The project investigates the video coding technologies for multiview video coding.

Computational Intelligence for Anomaly Detection in Networks: An Investigation
Description:  This project investigates use of computational intelligence (CI) for anomaly detection in computer networks.  Also, comparative performances of existing signature-based and misuse detection approaches are investigated.

Efficient Low-resolution Image Segmentation
Description:  The similarity measure criteria play a critical role in accurate segmentation.  This project aims at finding an appropriate similarity measure to extract objects in an image efficiently.

Agent Detections for Video Surveillance from Challenging Environment
Description:  This project investigates the strategies to identify active agents such as car, human, etc from a surveillance video sequence.

An analytical study of IT Security Governance (ITSG) in Australian organisations from international perspectives
Description:  This study will advance the IT security processes in Australian organistaions and compar it with international standards and guidelines for ITSG.

Understanding the movement of mobile nutrients in a drip irrigated vineyard with a clay loam soil
Description:

“Image Texture Segmentation Based on Wigner Distribution in a Fractional Fourier Domain
Description:  To apply the proposed WD-FrFT to image texture segmentation.  Since coarseness and directionality are two essential perceptual cues used by the human visual system for discriminating different textures, we will adopt two corresponding types of features, spatial - frequency and orientation, which will be extracted by L1 difference of Gaussian (DOG) filters and L2 wedge filters, respectively.

Difference equations and the problem of integrability
Description:  The evolution of a number of systems proceeds by steps and can be naturally described by difference equations.l  A special role in this class belongs to integrable equations.  Using tools like algebraic entropy and summetry analysis our aim is to better understand the behaviour of such systems and to develop efficient ways to solve them.

2D mesh refinement implementation in OpenFoam
Description:  The project verifies the accuracy of the two-dimensional adaptive mesh refinement for more than one refinements for quaedrilateral meshes by comparing with existing benchmark results.  Wide applications in practice will be followed after the successful accuracy verification of the adaptive mesh refinement method.

Statistical analysis of near infrared spectrometry
Description:  Near infrared (NIR) spectral analysis is a rapid, non-destructive assessment tool widely used to predict or describe quality parameters of material under investigation.  Important areas where this is applied include agriculture, food quality control, geography and geology, medical diagnostics, remote sending, oil, petrochemical and pharmaceutical industries.  the statisticl technoques of partial least squares or modified partial least squares are currently used in the calibration-prediction process, which relates the NIR spectra to the quality parameter/s.  The calibration process involves using a respresentative sample from the population under consideration, recording the NIR spectra and analysing the sample using usual reference methods, finally the calibration is used to define the relationship between these two measurements.  The current NIR spectral analysis procedures represent a 'black box', little understood by most practitioners, and the output is not easy to interpret.  This research project investigates the statistics underlying the present procedures in NIR analysis.  then we will seek more robust and transparent statistical procedures to replace or strengthen the current statistical practices.

Small area estimation of housing stress in Australia
Description:  Housing stress describes a financial situation of households where the cost of housing - either as rent, or as a mortgage repayment - is considered to be significantly high relative to household income.  It is also used to describe inadequate housing for a proportion of the population.  About 1.7 million people in Australia have housing stress.  Rises in housing prices and rents have generated keen interest in understanding who is struggling to afford to buy or rent a house, and what features exist at small area levels.  Housing stress statistics vary with demographic and socioeconomic conditions of households and with geography.  Due to the inavailability of micropopulation survey data in Australia, small area estimates for housing stress are not readily obtainable.  Policy makers need accurate housing stress estimates at local levels to make effective policy decisions on this and related issues such as poverty, social disadvantage, health and various social assistance programs.  This research project is concerned with providing these necessary small area housing stress estimates, using mocrosimulation modelling technollogy to simulate the characteristics of populations at the small area level.