Research in MLDA is shaping the future of automation and artificial intelligence, fusing theoretical and applied expertise to deliver transformative outcomes for industry and society.
Our research
Tactical behaviour discovery: real-time strategy games
In this project, researchers collaborated with Australia’s Defence Science and Technology organisation to develop a real-time strategy game agent. The novel development leveraged two machine-learning algorithms to outperform existing agents within the given environment in multiple areas.
PTZ-static camera for multiple objects tracking
Statics cameras have large Field-of-View (FoV), but objects tend to be of low resolution. PTZ cameras have a narrower field of view and subjects may appear larger, but they are also capable of mechanical pan, tilt and zoom operations. Our researchers have developed a set of novel algorithms and applications that use both PTZ and static cameras in tandem, harnessing the advantages of each for applications in real-time TV filming, such as in live sports.
Deep bypass: clear and dark real-time traffic profiling with deep learning
Working in collaboration with the Commonwealth Scientific and Industrial Research Organisation (CSIRO), our researchers have developed a deep learning-based model to detect malicious network attacks.
Predictive analytics to enhance Ausgrid's safety outcomes
Developed to enable Ausgrid to automatically identify factors related to safety incidents, this project involved a detailed analysis of multiple datasets to provide management insights around areas more likely to experience safety-related issues.
Optimising bus-bridge during train service interruptions
Our researchers developed several novel optimisation algorithms for Sydney Trains, utilising historical passenger data to automatically optimise existing and future bus replacement timetables. The project was primarily focused on improving the passenger experience.
MEET THE MINDS
behind our research excellence