..............................
...

RESEARCH

Our Expertise

  • Multi-Object Tracking (MOT): involves jointly estimating the number of objects and their trajectories from sensor data. In a multi-object system, the number of objects and their states vary with time due to objects appearing and disappearing. For example, the location, velocity and bearing of commercial planes at an airport, ships in a harbour, or pedestrians on the street. The objects are observed by a sensor (or sensors) such as radar, sonar, electro-optical, infrared, camera etc. Existing objects may not be detected and that false detections (or false alarm) may occur. As a result, at each time step the observation is a set of detections, only some of which are generated by existing objects and in general there is no information on which objects generated which detections. Driven by aerospace applications in the 1960’s, MOT has a long history spanning over 50 years. Today, MOT has found applications in diverse disciplines, including, air traffic control, surveillance, defence, space applications, oceanography, autonomous vehicles and robotics, remote sensing, computer vision, and biomedical research. MOT is combinatorial in complexity, and hence computationally intensive. State-of-the-arts MOT algorithms can simultaneously handle hundreds of objects and possibly in the thousands. The ISPL has recently developed a break-through solution, demonstrated to track over 1 million objects simultaneously, from approximately 1 billion detections.
  • Multi-Source Data Fusion: The goal of Data Fusion is to combine data from multiple heterogeneous sources, possibly in large volume, to construct more consistent, accurate, and useful information than that provided by any individual data source. The history of Data Fusion is deeply rooted in defence research where the aim is to use data from multiple sensors to provide better tracking of the targets of interest. Data Fusion has since evolved beyond defence into a plethora of applications from crowd monitoring, autonomous driving/vehicles, robotics, computer vision, to health and medical research etc. In general, any task that requires parameter estimation from multiple sources can benefit from Data Fusion. Data Fusion also employ a wide range of tools from probability/statistics to signal processing /control theory and artificial intelligence. It is still an area of focus in defence research in many countries including the US, China, UK, Australia, and the EU. The ISPL is a pioneer and world-leader in the random set approach to Data Fusion.
  • Optimisation: Apart from Convex Optimisation and its application to filter and filter bank design, speech enhancement as well as communications, we have specialised expertise in large-scale Multi-dimension Assignment problem. This combinatorial optimization problem is fundamental in the field of Operations Research with a host of applications such as data association, scheduling, resource allocation, e.g. assigning workers to jobs, machines to factory, marketing agents to regions, contracts to bidders; teachers to classes, and so on. Apart from the 2-D special case (also known as the linear assignment problem), Multi-dimensional Assignment is NP-hard. State-of-the-arts algorithm can accommodate large-scale problems that involves around 10 dimensions and 20 elements in each dimension. The ISPL has developed techniques that can solve problems with 100 dimensions and more than 80 elements per dimension.
  • Situational Awareness (SA): is one of the foundational tactical enablers (along with Command and Control, Information, and Intelligence) that permeate all aspects of Defence and National Security. SA refers to the perception of environmental elements and events with respect to space-time, the understanding of their meaning, and the projection of their future status region of responsibility. SA is vital not only for monitoring of potential threat in our air, sea and land space, but also for our personnel to make timely, intelligent judgements and provide appropriate responses in the event of a threat. The 2016 Australian Defence White Paper further highlights our Government’s effort to strengthen SA capabilities with substantial investment of Defence capital expenditure to boost Intelligence, Surveillance and Reconnaissance (ISR), electronic warfare, space, and cyber capabilities so that our forces have a comprehensive awareness of what is happening around them. ISPL specialises in large-scale solutions for SA, and automatic discovery/recognition of behavioural patterns.
  • Communications: Enabling technologies for communication in harsh environments is the key challenge in modern applications in wireless communications. There is a clear need for providing increased range and communication rate in harsh wireless environments such as shallow underwater and complex radio environments, as well as techniques for speech enhancement and noise suppression for voice communication. The ISPL focuses on statistical inference and optimisation methods that provide substantial improvement at reasonable computational costs as well as ease of implementation. Our underwater wireless communication team has a wealth of experience with multimodal underwater wireless communications, focusing on enhancing the rate, range, and reliability for various industry and defence applications, through innovative signal processing and communication algorithms. We have expertise in designing underwater acoustic and optical communication systems up to the prototype stage.
  • Signal Processing: The ISPL has a range of expertise in traditional Signal Processing areas ranging from blind signal separation, multi-rate signal processing, array processing, filter design, robust speech recognition, and assistive listening devices.

Start-up Companies

Nuheara www.nuheara.com  is a company established in early 2015 to bring to the market a Bluetooth connected augmented hearing device. That enables the user to use Bluetooth connected headphones without any attached cables to augment music and surround sound by using proprietary automatic mixing technology and intelligent audio processing to enable an improved communication experience in cafés and other social settings. It will also enable ease of use with portable devices for voice communication. The product IQbuds are selling online from Amazon and in major shops worldwide such as Best Buys, Target and Brookston in North America and Dixons in the UK. ISPL’s Professor Nordholm has been Nuheara’s Chief Scientist from the conception of the company until March 2018.

School of Electrical Engineering, Computing and Mathematical Sciences (EECMS) | Faculty of Science and EngineeringSchool of EECMS