Joint System Identification for Point Processes and Time-series. In various application areas such as neurophysiology, earthquake modeling, price spikes in electricity markets, the data of interest are point processes (aka sequences of events) or combinations of point processes and analog signals. To understand the underlying subject of interest we need to be able to extract the maximum information from these observation sequences. The current tools for doing this are very limited. This resear ....Joint System Identification for Point Processes and Time-series. In various application areas such as neurophysiology, earthquake modeling, price spikes in electricity markets, the data of interest are point processes (aka sequences of events) or combinations of point processes and analog signals. To understand the underlying subject of interest we need to be able to extract the maximum information from these observation sequences. The current tools for doing this are very limited. This research program will develop the complex signal processing and system methodology needed to create a suitable tool set.Read moreRead less
Discovery Early Career Researcher Award - Grant ID: DE120102388
Funder
Australian Research Council
Funding Amount
$375,000.00
Summary
From Bayesian filtering to smoothing and prediction for multiple object systems. This project will develop new and improved algorithms for tracking multiple targets, such as tanks, submarines or planes, using the state of the art in mathematical and computational design. These will enable more efficient and accurate technologies for defence related applications including intelligence, surveillance and reconnaissance.
Parameter estimation for multi-object systems. Parameter estimation in multi-object system is essential to the application of multi-object filtering to a wider range of practical problems with social and commercial benefits. This project develops the necessary parameter estimation techniques for complete 'plug-and-play' multi-object filtering solutions that facilitates widespread applications.
A Bayesian framework for frequency domain identification. The national and social benefits of the project will be reflected
through the application recognition of the research work in the various industry and research community; and also through our international collaboration. The national and social benefits are also delivered by producing specialized researchers and engineers in systems and control engineering. These people include the research students who will participate in and learn f ....A Bayesian framework for frequency domain identification. The national and social benefits of the project will be reflected
through the application recognition of the research work in the various industry and research community; and also through our international collaboration. The national and social benefits are also delivered by producing specialized researchers and engineers in systems and control engineering. These people include the research students who will participate in and learn from the proposed project.Read moreRead less
A stochastic geometric framework for Bayesian sensor array processing. This project develops a mathematical framework, and a new generation of techniques, for sensor array processing to address real-world problems with uncertainty in array parameters and number of signals. The outcomes will enhance the capability of sensors in many application areas including, radar, sonar, astronomy and wireless communications.
Point processes system identification under simultaneity. Neuroscientists study neuronal brain dynamics of mammals via recordings from scores of tiny electrodes. But analysing these experiments is a problem because current methods cannot handle the common case where neurons discharge simultaneously. This project aims to provide powerful new tools to overcome this bottleneck.
Riemannian System Identification. A growing number of applications such as satellite attitude estimation, pose estimation in computer vision and direction estimation in statistics require the study of random processes in Riemannian manifolds and Lie Groups. This project will provide: methods for the construction/ numerical simulation of such processes; methods of system identification and their asymptotic performance analysis; and, algorithms for process state estimation.
Modeling stochastic systems in Riemannian manifolds. This project aims to develop new statistical signal processing and control engineering algorithms and tools that will enable tracking of objects remotely on land and in space. A growing number of applications require the study of random processes in Riemannian manifolds, that is processes that evolve subject to a geometric constraint. This project aims to provide methods for the numerical simulation of such processes, methods of online and off ....Modeling stochastic systems in Riemannian manifolds. This project aims to develop new statistical signal processing and control engineering algorithms and tools that will enable tracking of objects remotely on land and in space. A growing number of applications require the study of random processes in Riemannian manifolds, that is processes that evolve subject to a geometric constraint. This project aims to provide methods for the numerical simulation of such processes, methods of online and offline system identification from data on such processes and asymptotic performance analysis; and algorithms for process state estimation that obeys the geometry. The outcomes will advance and benefit spatio-temporal data analysis in areas such as transport, health and video-security.Read moreRead less
The Time-Varying Eigenvalue Problem with Application to Signal Processing and Control. Linear models are ubiquitous in representing physical processes. Decomposing a linear model into its fundamental components is known as the eigenvalue problem. In applications as wide ranging as astronomy, aircraft control systems, Internet search engines and communication systems, it is necessary to perform this decomposition of a pertinent time varying linear model on the fly. This project aims to develop si ....The Time-Varying Eigenvalue Problem with Application to Signal Processing and Control. Linear models are ubiquitous in representing physical processes. Decomposing a linear model into its fundamental components is known as the eigenvalue problem. In applications as wide ranging as astronomy, aircraft control systems, Internet search engines and communication systems, it is necessary to perform this decomposition of a pertinent time varying linear model on the fly. This project aims to develop significantly faster and more accurate algorithms for this time varying eigenvalue problem than currently exist. Very modern techniques will be employed to achieve this aim, and the potential benefits to Australian hi-tech industries are great.
Read moreRead less
Automated vision-based aircraft collision warning technologies. Australia is a sparsely populated country with a number of unique airspace features. This project will investigate novel vision-based collision warning systems that can improve safety for piloted aircraft and also help achieve integration of UASs (Uninhabited Aerial Systems) into national airspace. The benefits of UAS technologies are particularly relevant to Australia, as governments and industry struggle to cope with providing equ ....Automated vision-based aircraft collision warning technologies. Australia is a sparsely populated country with a number of unique airspace features. This project will investigate novel vision-based collision warning systems that can improve safety for piloted aircraft and also help achieve integration of UASs (Uninhabited Aerial Systems) into national airspace. The benefits of UAS technologies are particularly relevant to Australia, as governments and industry struggle to cope with providing equivalent levels of service to remote communities over vast distances (or border protection of vast regions). The population base of Australia requires that cost-effective solutions are sought to meet this end. Read moreRead less