Projects

Project Leader: Dr Antonio Robles-Kelly
Reference Customers/Collaborators: Canon Information Systems Research Australia Pty Ltd (CISRA)
Key competitive advantages: accurate signature identification that is robust to multiple light sources and highly textured surfaces, superior shape identification and recognition that eliminates shadowing and polarimetric effects, faster signature and shape identification, less end-user intervention required, strong patent position NICTA is developing imaging spectroscopy tools to provided enriched scene understanding. This will involve understanding an object’s composition and the concentration of its material components. Current technologies use spatial techniques (such as shape understanding) based on absorption classification methods to understand scenes. NICTA combines spatial techniques with a statistical approach based on reflection classification methods to understand scenes.

Automated Pattern Recognition Technology

Project Leader: Dr Yongsheng Gao
Reference Customers/Collaborators: Depart of Agriculture, Fisheries and Forestry, Biosecurity Queensland Control Centre
Key competitive advantages: in situ monitoring and identification/interpretation
NICTA has developed a field microscope for the collection of images on a laptop or PDA. NICTA researchers are also developing software utilising automated pattern recognition algorithms to automatically identify insect species.

Project Leader: Prof Gernot Heiser
NICTA will develop key secure virtualisation technologies and architectural analysis techniques to assist in the delivery of systems with guaranteed secure information handling. NICTA will apply in-house tools, techniques and systems in contexts where true dependability is important, and hard or impossible to achieve with other means. The team is developing tractible and affordable design solutions, including architectures and techniques to identify the critical elements that require verification, and increased automation of verification tasks.

Project Leader: Dr Jon Gray
NICTA has extensive experience in building tools to support policy and service delivery agencies in project-level and enterprise level planning, design, and implementation, including the development of Service Oriented Architectures (SOA's); architecture performance assessment and capacity planning; Integration and interoperability of SOA with legacy systems and cost and effort estimation of major re-architecting projects.

Optimisation

Project Leaders: Dr Phil Kilby, Dr Toby Walsh
NICTA has expertise in combinatorial optimisation (ie where there are many decision variables each having a finite number of choices) NICTA has expertise in solving problems in diagnosis and planning, packing; routing and scheduling and problems involving 2D placement (eg. diagrams) and 3D structure (eg protein structure).

Project Leaders: Prof Rob Evans, Prof Stan Skafidas, Prof Bill Moran
NICTA has expertise in radar systems, target tracking, radar signal processing and waveform design. In addition the team has expertise in microelectronics development particularly in RFCMOS. NICTA has developed a low cost flexible adaptive digital radar based silicon technology with integrated processing algorithms. Operating at 76-77GHz frequency band, it supports both long and short range radar application requirements. The team also has significant expertise in the 802.11ad standard and antennas for that standard as well as in very high speed analog-to-digital converters, which can be used to enable the design of communication receivers with very high bandwidths.

Project Leader: Dr Fang Chen
NICTA is developing tools for automatic analysis of cognitive load for individuals undertaking work tasks. The current version is primarily based on speech features, which is the most mature methodology. The team also has some lower maturity methods including eye movement, pen pressure and skin conductance. The tools may be used in real-time or off-line and can be applied to the recruitment of staff, development of training programs or monitoring of staff for cognitive overload in real-time work situations.

Project Leader: Prof Brian Lovell
NICTA has a number of real-time face identification and tracking algorithms that enable a person to be identified for surveillance purposes. The NICTA technologies feature low memory usage, accuracy even with non-frontal poses and scaleability (it can work with multiple processors and CPUs with very large databases and in parallel it can work with large scale systems with hundreds and thousands of cameras) processing over two million faces in under a second.

Video Analysis for Surveillance

Project Leader: Dr Jian Zhang
NICTA has developed methods for automated analysis of video scenes for thermal and optical images. In particular, the team is experienced in detection of vehicles, people and small vessels (thermal only). Researchers also have expertise in detecting human activity and movement patterns.

Project Leader: Dr Marius Portmann
NICTA has experience in the research and design of wireless mesh routing protocols and wireless mesh performance testing. In particular, the mesh routing protocols address the need of mesh networks to adapt to environmental operating conditions such as the particular placement of the nodes or the presence of interference.

Forensic Informatics

Project Leader: Dr Jason Tangen
NICTA is currently reviewing the current visualisation techniques, methods, and decision support tools used by forensic identification specialists. They are also• measuring bias and accuracy for normal observers as they engage with these decision making tools. It is expected that the outcomes will inform both new evidence-based technology developments and training techniques for forensic professionals.

Project Leader: Prof Brian Anderson
NICTA has globally convergent algorithms for distributed control of multiple-vehicle systems such as groups of autonomous aerial vehicles; wireless sensor network localization algorithms; effective data fusion methods capable of handling non-standard measurements such as natural language; signal processing methods to work out the locations of individual emitters from a combination of multiple individual received spectra; reliable methods of communication and message passing in the face of interference or mobility; secure design methods for networked estimation and control in order to safeguard critical infrastructure.