Human models for accelerated robot learning and human-robot interaction. This project aims to develop novel approaches to teach robots to proficiently interact with humans in a safe and low-cost manner. To achieve this aim, this project will develop novel models from which various human behaviours can be generated and used to train human-robot interaction policies in simulation. Expected outcomes of this project include new computational models of human behaviour built using cognitive science th ....Human models for accelerated robot learning and human-robot interaction. This project aims to develop novel approaches to teach robots to proficiently interact with humans in a safe and low-cost manner. To achieve this aim, this project will develop novel models from which various human behaviours can be generated and used to train human-robot interaction policies in simulation. Expected outcomes of this project include new computational models of human behaviour built using cognitive science theories and limited data and new training schemes for robot learning in simulation. By training robots in simulation with accurate human models, this research will enable fast and safe robot training to support the deployment and adoption of robots in human contexts such as healthcare facilities, homes, and workplaces.Read moreRead less
Constraint-based Reasoning for Multi-agent Pathfinding. Automation is a transformative technology for logistics -- using robots to manipulate inventory allows warehouses to be more efficient, and larger-scale, than ever before. But doing this in practice requires efficient, reliable methods for coordinating ever-larger fleets of robots. These problems are extremely difficult, and current approaches either scale poorly or give weak or no guarantees on solution quality. The project will develop t ....Constraint-based Reasoning for Multi-agent Pathfinding. Automation is a transformative technology for logistics -- using robots to manipulate inventory allows warehouses to be more efficient, and larger-scale, than ever before. But doing this in practice requires efficient, reliable methods for coordinating ever-larger fleets of robots. These problems are extremely difficult, and current approaches either scale poorly or give weak or no guarantees on solution quality. The project will develop transformative approaches to multi-agent pathfinding which can handle industrial size problems, and handle all of the complications that arise in practical applications. This will deliver improved cost-effectiveness and productivity to automated warehouse logistics and other agent coordination problems.Read moreRead less
Searching for near-exact protein models. This project aims to develop novel and efficient heuristic-based algorithms leading to near accurate protein tertiary structure models. Knowledge about protein structures is fundamental to our understanding of living systems. The progress on experimental determination of these structures has been extremely limited and remains an open challenge in molecular biology. Computational prediction of protein structures from sequences is emerging as a promising ap ....Searching for near-exact protein models. This project aims to develop novel and efficient heuristic-based algorithms leading to near accurate protein tertiary structure models. Knowledge about protein structures is fundamental to our understanding of living systems. The progress on experimental determination of these structures has been extremely limited and remains an open challenge in molecular biology. Computational prediction of protein structures from sequences is emerging as a promising approach, but its accuracy is far from satisfactory. The software systems developed in this project will be used in structural identification of target proteins in drug design. This will make drug design process more efficient, saving time and cost, potentially saving lives.Read moreRead less
Small Scalable Natural Language Models using Explicit Memory. Deep neural networks have had spectacular success in natural language processing, seeing wide-spread deployment as part of automatic assistant devices in homes and cars, and across many valuable industries including finance, medicine and law. Fueling this success is the use of ever larger models, with exponentially increasing training resources, accompanying hardware and energy demands. This project aims to develop more compact models ....Small Scalable Natural Language Models using Explicit Memory. Deep neural networks have had spectacular success in natural language processing, seeing wide-spread deployment as part of automatic assistant devices in homes and cars, and across many valuable industries including finance, medicine and law. Fueling this success is the use of ever larger models, with exponentially increasing training resources, accompanying hardware and energy demands. This project aims to develop more compact models, based on the incorporation of an explicit searchable memory, which will dramatically reduce model size, hardware requirements and energy usage. This will make modern natural language processing more accessible, while also providing greater flexibility, allowing for more adaptable and portable technologies.Read moreRead less
Efficient and effective methods for classifying massive time series data. This project aims to transform the theory and practice of time series classification. The current state of the art cannot handle the massive numbers of time series that describe many critical problems facing humanity, such as disease transmission and climate change. This project seeks to develop methods that can analyse dynamic processes at global scale, delivering the most accurate classifiers feasible within a given comp ....Efficient and effective methods for classifying massive time series data. This project aims to transform the theory and practice of time series classification. The current state of the art cannot handle the massive numbers of time series that describe many critical problems facing humanity, such as disease transmission and climate change. This project seeks to develop methods that can analyse dynamic processes at global scale, delivering the most accurate classifiers feasible within a given computational budget. Expected outcomes of this project include efficient, effective and broadly applicable time series classification technologies. This should provide significant benefits to myriad sectors, transforming data science for time series problems and supporting innovation in industry, commerce and government.Read moreRead less
Motion of objects in soils. This project aims to conduct a fundamental study of a challenging class of geotechnical problems in which an object moves inside a layer of soil, interacts with soil, and disturbs it, by developing advanced numerical and analytical methods. This project expects to determine the fundamental principles governing soil behaviour upon movement of embedded objects. The expected outcomes are robust solutions and computational procedures that will benefit government and engin ....Motion of objects in soils. This project aims to conduct a fundamental study of a challenging class of geotechnical problems in which an object moves inside a layer of soil, interacts with soil, and disturbs it, by developing advanced numerical and analytical methods. This project expects to determine the fundamental principles governing soil behaviour upon movement of embedded objects. The expected outcomes are robust solutions and computational procedures that will benefit government and engineers by providing safer and more cost-effective strategies for designing, constructing, and maintaining Australia's infrastructure. This should bring significant benefits to industries engaged in harvesting energy resources, such as wind farms, as well as oil and gas.Read moreRead less
Explainable Artificial Creativity. This project aims to develop explainable models for creative AI systems which enable more productive and satisfying interactions between them and their human co-creators. This will boost both human and machine creativity through sustained, ongoing exchanges, leading to high-quality creative outcomes via automated ideation and more advanced human-machine collaborations. The proposed techniques will be validated with creative professionals, ensuring practical ind ....Explainable Artificial Creativity. This project aims to develop explainable models for creative AI systems which enable more productive and satisfying interactions between them and their human co-creators. This will boost both human and machine creativity through sustained, ongoing exchanges, leading to high-quality creative outcomes via automated ideation and more advanced human-machine collaborations. The proposed techniques will be validated with creative professionals, ensuring practical industry relevance. We expect the outcomes to include new methods that automatically generate persuasive explanations, new forms of communication including dialogues between creative AI systems and users, and new understanding of general aspects of explainability for creative AI systems.Read moreRead less
Advanced Machine Learning with Bilevel Optimization. There is an urgent need to develop a new machine learning (ML) paradigm that can overcome data-privacy and model-size constraints in real-world applications. This project aims to develop an advanced paradigm of ML with bilevel optimisation, called bilevel ML. A theoretically-guaranteed fast approximate solver and a new fuzzy bilevel learning framework will be developed to achieve the aim in complex situations; a methodology to transfer knowled ....Advanced Machine Learning with Bilevel Optimization. There is an urgent need to develop a new machine learning (ML) paradigm that can overcome data-privacy and model-size constraints in real-world applications. This project aims to develop an advanced paradigm of ML with bilevel optimisation, called bilevel ML. A theoretically-guaranteed fast approximate solver and a new fuzzy bilevel learning framework will be developed to achieve the aim in complex situations; a methodology to transfer knowledge and an approach to fast-adapt bilevel optimization solutions when required computing resources change. The anticipated outcomes should significantly improve the reliability of ML with benefits for safety learning and computing resource optimisation in ML-based data analytics.Read moreRead less
In for the count: Maximising trust and reliability in Australian elections. This project aims to develop innovative approaches to identifying, measuring, and evaluating errors and purposeful intervention in the uniquely complex elections at the basis of Australian democracy. Such methods can underpin a world-class election auditing system, which contends with the risks that are emerging at the intersection of election digitisation, cybersecurity and foreign interference. The project’s expected o ....In for the count: Maximising trust and reliability in Australian elections. This project aims to develop innovative approaches to identifying, measuring, and evaluating errors and purposeful intervention in the uniquely complex elections at the basis of Australian democracy. Such methods can underpin a world-class election auditing system, which contends with the risks that are emerging at the intersection of election digitisation, cybersecurity and foreign interference. The project’s expected outcomes are new auditing methods, tested on real Australian election data, with their benefits quantified against global best practice. The research outputs should help reinforce the community’s trust in Australian elections, which are a foundation for our security, social cohesion, and political resilience.Read moreRead less
Fairness in Natural Language Processing. Natural language processing (NLP) has achieved spectacular commercial successes in recent years, and has been deployed across an ever-increasing breadth of devices and application areas. At the same time, there has been stark evidence to indicate that naively-trained models amplify biases in training data, and perform inconsistently across text relating to different demographic groupings of individuals. This project aims to systematically quantify the ext ....Fairness in Natural Language Processing. Natural language processing (NLP) has achieved spectacular commercial successes in recent years, and has been deployed across an ever-increasing breadth of devices and application areas. At the same time, there has been stark evidence to indicate that naively-trained models amplify biases in training data, and perform inconsistently across text relating to different demographic groupings of individuals. This project aims to systematically quantify the extent of such biases, and develop models that are both more socially equitable, as well as less prone to expose private data in the learned representations. In doing so, it will make NLP more accessible to new populations of users, and remove socio-technological barriers to NLP uptake.Read moreRead less