Multi-scale recognition: generating meaning from multi-resolution data. The next generation of robots will be able to precisely recognise objects to reason about the world. This project will develop robust recognition systems that will aid robots in providing assistance in populated urban areas as well as in monitoring underwater environments for marine biodiversity preservation.
Automated benthic understanding with multimodal observations. This project aims to deliver cost-effective techniques to explore and monitor marine environments. The project will develop novel methods for classification of large extent, multimodality seafloor surveys consisting of high-resolution visual 3D gigamosaics made of tens of thousands of images coregistered with broad-scale, lower resolution remote sensing data. This knowledge is essential for designing cost-effective, scalable systems t ....Automated benthic understanding with multimodal observations. This project aims to deliver cost-effective techniques to explore and monitor marine environments. The project will develop novel methods for classification of large extent, multimodality seafloor surveys consisting of high-resolution visual 3D gigamosaics made of tens of thousands of images coregistered with broad-scale, lower resolution remote sensing data. This knowledge is essential for designing cost-effective, scalable systems to explore, map and monitor Australia's marine environments. At a broader level, the approach and the techniques developed in this project have the potential to have applications in other areas such as terrestrial and intertidal ecology, extending positive impacts beyond benthic environments.Read moreRead less
Robust learning of dynamic systems. Robots and other autonomous machines use models of the real world to predict the result of their actions and make decisions, but existing methods used for machine-learning are unreliable in many cases and can be easily fooled. This project aims to make machine-learning of dynamic system models reliable, accurate, and secure. The outcomes of this project will be new models and algorithms that ensure safety and increase accuracy of models learned from data. This ....Robust learning of dynamic systems. Robots and other autonomous machines use models of the real world to predict the result of their actions and make decisions, but existing methods used for machine-learning are unreliable in many cases and can be easily fooled. This project aims to make machine-learning of dynamic system models reliable, accurate, and secure. The outcomes of this project will be new models and algorithms that ensure safety and increase accuracy of models learned from data. This project will benefit robotics, control engineering, infrastructure automation, and other fields that demand the capability to model physical systems from limited data. It will also improve cybersecurity by making learning algorithms resilient to deliberate attacks with false data.Read moreRead less
Discovery Early Career Researcher Award - Grant ID: DE120103051
Funder
Australian Research Council
Funding Amount
$375,000.00
Summary
Data fusion and active sensing for environment monitoring. This project aims to create a novel statistical framework for data fusion that integrates elements of perception and machine learning for better understanding of natural phenomena. The outcome will be a methodology for the seamless integration of space-time correlated data that will revolutionise the use of multi-model information.