Defense against adversarial attacks on deep learning in computer vision. Computer vision applications rely heavily on deep learning, which is highly vulnerable to being fooled by adding subtle perturbations to object/image textures that are imperceptible to humans. This project aims to develop defense mechanisms to detect and remove adversarial patterns from the input images. The project expects to advance knowledge in understanding the vulnerabilities of deep learning, and to design deep learni ....Defense against adversarial attacks on deep learning in computer vision. Computer vision applications rely heavily on deep learning, which is highly vulnerable to being fooled by adding subtle perturbations to object/image textures that are imperceptible to humans. This project aims to develop defense mechanisms to detect and remove adversarial patterns from the input images. The project expects to advance knowledge in understanding the vulnerabilities of deep learning, and to design deep learning architectures that are inherently robust. The outcomes of this project will increase the security and reliability of computer vision by detecting, reporting and nullifying such attacks and will benefit the general public and industry on many fronts.Read moreRead less
“Beacons in the Night” unveiling how galaxies light up dark matter. How dark matter influences the formation and evolution of galaxies is to this day an outstanding question in astrophysics. To answer it, world-class facilities and a unique combination of observations and theory are required. This DP team, a world-class team of observers and theorists, will tackle this question by leveraging on two multi-million dollar projects: the MAGPI galaxy survey and the hydrodynamical simulations suite EA ....“Beacons in the Night” unveiling how galaxies light up dark matter. How dark matter influences the formation and evolution of galaxies is to this day an outstanding question in astrophysics. To answer it, world-class facilities and a unique combination of observations and theory are required. This DP team, a world-class team of observers and theorists, will tackle this question by leveraging on two multi-million dollar projects: the MAGPI galaxy survey and the hydrodynamical simulations suite EAGLE-XL. MAGPI will deliver exquisite kinematics for hundreds of galaxies in the middle ages of the Universe, providing a view to the effect of dark matter on galaxies at this critical time, while EAGLE-XL represents the technological frontier in simulations and provides the best interpretative framework for MAGPI.Read moreRead less
Using immersive virtual reality to enhance students’ science visualisation. This project aims to investigate the potential of advanced visualisation technology, immersive virtual reality, as a collaborative learning environment to support students to explore their ideas and learn a core chemistry concept, molecular structures and functions. Incorporating both data analytics and qualitative video analysis, the project will establish a deep understanding of how students learn about the molecular w ....Using immersive virtual reality to enhance students’ science visualisation. This project aims to investigate the potential of advanced visualisation technology, immersive virtual reality, as a collaborative learning environment to support students to explore their ideas and learn a core chemistry concept, molecular structures and functions. Incorporating both data analytics and qualitative video analysis, the project will establish a deep understanding of how students learn about the molecular world within an immersive virtual reality environment in relation to the learners’ experience, chemistry content, visual representations, learning tasks, and design features. The project will recommend evidence-driven design considerations for learning resources development and further educational research with advanced technologies.Read moreRead less
Fine-grained Human Action Recognition with Deep Graph Neural Networks. This project aims to develop novel graph neural network based deep learning algorithms for fine-grained human action recognition. This project expects to bring human action analysis to the next level and to significantly advance the analysis of subtle yet complex human actions. Expected outcomes of this project include theoretical advances on graph representation based deep learning algorithms for spatial-temporal data, and e ....Fine-grained Human Action Recognition with Deep Graph Neural Networks. This project aims to develop novel graph neural network based deep learning algorithms for fine-grained human action recognition. This project expects to bring human action analysis to the next level and to significantly advance the analysis of subtle yet complex human actions. Expected outcomes of this project include theoretical advances on graph representation based deep learning algorithms for spatial-temporal data, and enabling techniques for more objective human action analysis in many domains such as sports and health. This should provide significant benefits to any application domain involving big and complex spatial-temporal data for finer analytics and better knowledge discovery.Read moreRead less
Going wild: Neural processing in freely moving animals. This project aims to use new techniques in wireless neural recording to reveal how small neural networks process visual information to make fast, accurate decisions. The project is designed to generate new knowledge about biological solutions to contextual information processing and how tiny, simple biological neural systems control critical animal behaviours such as predator avoidance. Expected outcomes will be new biological insights with ....Going wild: Neural processing in freely moving animals. This project aims to use new techniques in wireless neural recording to reveal how small neural networks process visual information to make fast, accurate decisions. The project is designed to generate new knowledge about biological solutions to contextual information processing and how tiny, simple biological neural systems control critical animal behaviours such as predator avoidance. Expected outcomes will be new biological insights with which to develop novel bio-inspired decision-making processing systems as required in small, autonomous robots. The anticipated benefits of this project will be advances in fundamental neuroscience and animal behaviour and is expected to provide significant value to a fast-developing industrial sector.Read moreRead less
Complexity of group algorithms and statistical fingerprints of groups. This project aims to shape the next generation of efficient randomised algorithms in the field of group theory, the mathematics of symmetry. Fundamental mathematics underpins modern technological tasks such as web searches, sorting and data compression. This project aims to determine characteristic statistical fingerprints of key building-block groups. These group statistics lead to much faster procedures to essentially facto ....Complexity of group algorithms and statistical fingerprints of groups. This project aims to shape the next generation of efficient randomised algorithms in the field of group theory, the mathematics of symmetry. Fundamental mathematics underpins modern technological tasks such as web searches, sorting and data compression. This project aims to determine characteristic statistical fingerprints of key building-block groups. These group statistics lead to much faster procedures to essentially factor huge groups into smaller building-block groups in a manner akin to factoring an integer into its prime factors. The anticipated goal is to include the outcomes in publicly available symbolic algebra computer packages. As the theory of symmetry has broad applications in the mathematical and physical sciences, there is the potential for far reaching benefits.Read moreRead less
Advancing programmable genetic computation to control plant gene activity. Plants can sense diverse internal and external conditions and integrate them to appropriately tune their response and maximize fitness. Plant biotechnology relies heavily on manipulating gene activity to change cell functions and confer advantageous agronomic traits. However, our ability to control plant gene activity remains rudimentary, limiting our biotechnology capabilities. This project aims to develop synthetic gene ....Advancing programmable genetic computation to control plant gene activity. Plants can sense diverse internal and external conditions and integrate them to appropriately tune their response and maximize fitness. Plant biotechnology relies heavily on manipulating gene activity to change cell functions and confer advantageous agronomic traits. However, our ability to control plant gene activity remains rudimentary, limiting our biotechnology capabilities. This project aims to develop synthetic gene logic gates in plants, to enable the construction of programmable genetically-encoded computational functions that can sense and process customizable inputs to drive desired changes in plant function. This advance will underpin useful applications in plant biotechnology such as improved crop stress tolerance and yield.Read moreRead less
Advancing the visualisation and quantification of nephrons with MRI. . This project aims to characterise key components of nephrons, the glomeruli and tubules, using magnetic resonance imaging without contrast agents, in combination with Deep Learning and super-resolution techniques. Nephrons, the basic functional unit of the kidney, are critical to the maintenance of the body’s homeostasis. Their number and architecture are critical determinants of kidney function. The expected outcomes are inn ....Advancing the visualisation and quantification of nephrons with MRI. . This project aims to characterise key components of nephrons, the glomeruli and tubules, using magnetic resonance imaging without contrast agents, in combination with Deep Learning and super-resolution techniques. Nephrons, the basic functional unit of the kidney, are critical to the maintenance of the body’s homeostasis. Their number and architecture are critical determinants of kidney function. The expected outcomes are innovative semi-automated nephron visualisation and quantitation tools that enable efficient renal phenotyping. Techniques tailored to widely accessible preclinical research scanners are expected to accelerate research into genetic and environmental factors affecting kidney microstructure in embryonic and post-natal life.Read moreRead less
The roles and regulators of new plant cells linked to root transport. Plant genomics has moved to the single cell resolution, allowing precise investigations of previously hidden cell types and cell states that respond to environmental stress and that vary among differentially adapted plant populations. Here, we will extend our pioneering efforts that have mapped and discovered novel root cell types, to determine their salt and nutrient stress responses, and to elegantly dissect the underling ca ....The roles and regulators of new plant cells linked to root transport. Plant genomics has moved to the single cell resolution, allowing precise investigations of previously hidden cell types and cell states that respond to environmental stress and that vary among differentially adapted plant populations. Here, we will extend our pioneering efforts that have mapped and discovered novel root cell types, to determine their salt and nutrient stress responses, and to elegantly dissect the underling causal genetic variation. The unique cell markers and regulatory networks will be validated with tissue specific and transgenic tools that can work across a host of plant species to reveal adaptive cellular responses to harsh environmental conditions.Read moreRead less