Keynote Speakers
Please check back regularly as more information will be added as it is received from each speaker.
Fang Chen
Distinguished Professor Fang Chen is a globally recognized leader in AI and data science. She currently serves as the Executive Director of the Data Science Institute at the University of Technology Sydney. Previously, she held roles as Dean of the Faculty at Beijing Jiaotong University and senior leadership positions at Intel, Motorola, and the Commonwealth Scientific and Industrial Research Organisation (CSIRO). She serves as the Steering Committee Chair for ACM Intelligent User Interfaces.
Her extensive expertise lies in developing innovative, data-driven solutions that address complex challenges across large-scale networks in sectors such as transportation, water, energy, agriculture, telecommunications, education, health, and real estate. She is a committed advocate for ethical and human-centered AI practices.
Talk Title: Advancing Human Potential Through AI
In a time of rapid technological evolution, the intersection of artificial intelligence and human potential offers remarkable opportunities for advancing experiences, expertise, training, and performance. This keynote will delve into how AI-driven systems can be harnessed to optimize human-machine teaming, fostering environments where humans and machines work seamlessly to enhance outcomes.
Our work addresses key challenges at the intersection of human and machine interactions. By integrating methodologies from AI, data science, human-computer interaction (HCI), and behavioural sciences and neurosciences, we seek to understand and positively influence human behaviour and interaction with information, systems, robots, and each other. I will cover key topics of AI-driven multimodal cognitive load measurement, human-machine trust and calibration, adaptive learning for humans and AI, and the ethical implications of these advancements. We will explore adaptive learning systems, personalized programs, and intelligent solutions that augment human performance, emphasizing practical applications and case studies.
Finally, the talk will address the ethical considerations and challenges of implementing AI systems that align with human values, aiming for a future where AI is a transformative force for realizing human potential in a dynamic world.”
Nikola K. Kasabov
Professor Nikola K Kasabov is a Life Fellow of IEEE, Fellow of the Royal Society of New Zealand, Fellow of the INNS College of Fellows, DVF of the Royal Academy of Engineering UK. He has Doctor Honoris Causa from Obuda University, Budapest. He is the Founding Director of KEDRI and Professor at Auckland University of Technology and director of Knowledge Engineering Ltd. He is also Visiting Professor at IICT Bulgarian Academy of Sciences and Dalian University, China and Honorary Professor at the University of Auckland NZ and Peking University in Shenzhen.
Talk Title: Evolving Multimodal Associative Memories in Brains and Machines: This is all we need
Evolving multimodal associative memories (EMAM) are systems that associate and capture incrementally and continuously related items, objects and processes of multiple modalities to create dynamic structures in time and space, that can be recalled/triggered using partial modality information. Examples are the spatio-temporal associative memories in brains and nature, where both spatial and temporal information are dynamically integrated. Despite the fact that majority of data, that is dealt with in machine learning and AI across information and data sciences, are multimodal temporal- or spatio/spectro temporal, there are still no efficient methods for building EMAM. Early methods, such as Hopfield networks and Kosko’s bidirectional associative memories, still very popular nowadays, are designed to deal with static, vector-based data. The now popular large language models can generate associations specifically for semantic entities based on pre-trained networks.
The talk argues that the human brain functions in its cognitive functions, including consciousness, as a dynamic spatio-temporal EMAM. Then the talk introduces brain-inspired spiking neural network (SNN) architectures, exemplified by NeuCube [1,2] for the development of EMAM. Applications for classification and prediction of biological and brain signals, audio-visual data, environmental data, financial and economic data are presented.
Future directions are outlined towards the development of hybrid SNN, where a SNN is used for capturing spatio-temporal characteristics from continusly incoming data and other ML methods are used for evolving classification, prediction and knowledge discovery. The talk concludes that EMAM could be the way for the future brain-inspired AI [3], including conscious machines.
Danilo Mandic
Danilo P. Mandic is a Professor of Machine Intelligence with Imperial College London, UK, and has been working in the areas of machine intelligence, statistical signal processing, big data, data analytics on graphs, bioengineering, and financial modelling. He is a Fellow of the IEEE and the current President of the International Neural Network Society (INNS). Dr Mandic is the Director of the Laboratory for Artificial Intelligence and Data Analytics (AIDA-LAb, www.aidalab.co.uk), and has more than 600 publications in international journals and conferences. He has published two research monographs on neural networks, entitled “Recurrent Neural Networks for Prediction”, Wiley 2001, and “Complex Valued Nonlinear Adaptive Filters: Noncircularity, Widely Linear and Neural models”, Wiley 2009 (both first books in their respective areas), and has co-edited books on Data Fusion (Springer 2008) and Neuro- and Bio-Informatics (Springer 2012).
Talk Title: Interpretable Convolutional NNs and Graph CNNs: Role of Domain Knowledge
The success of deep learning (DL) and convolutional neural networks (CNN) has also highlighted that NN-based analysis of signals and images of large sizes poses a considerable challenge, as the number of NN weights increases exponentially with data volume – the so called Curse of Dimensionality. In addition, the largely ad-hoc fashion of their development, albeit one reason for their rapid success, has also brought to light the intrinsic limitations of CNNs - in particular those related to their black box nature. To this end, we revisit the operation of CNNs from first principles and show that their key component – the convolutional layer – effectively performs matched filtering of its inputs with a set of templates (filters, kernels) of interest. This serves as a vehicle to establish a compact matched filtering perspective of the whole convolution-activation-pooling chain, which allows for a theoretically well founded and physically meaningful insight into the overall operation of CNNs. This is shown to help mitigate their interpretability and explainability issues, together with providing intuition for further developments and novel physically meaningful ways of their initialisation. Such an approach is next extended to Graph CNNs (GCNNs), which benefit from the universal function approximation property of NNs, pattern matching inherent to CNNs, and the ability of graphs to operate on nonlinear domains. GCNNs are revisited starting from the notion of a system on a graph, which serves to establish a matched-filtering interpretation of the whole convolution-activation-pooling chain within GCNNs, while inheriting the rigour and intuition from signal detection theory. This both sheds new light onto the otherwise black box approach to GCNNs and provides well-motivated and physically meaningful interpretation at every step of the operation and adaptation of GCNNs. It is our hope that the incorporation of domain knowledge, which is central to this approach, will help demystify CNNs and GCNNs, together with establishing a common language between the diverse communities working on Deep Learning and opening novel avenues for their further development.
Giovanni Russello
Professor Giovanni Russello is a Cyber Security specialist and the Head of the School of Computer Science at the University of Auckland, New Zealand. Giovanni previously served as the Director of the Cyber Security Research Programme, a multi-million-dollar project funded by the Ministry of Business Innovation and Enterprise aimed at enhancing New Zealand's cyber security posture and fostering collaboration between New Zealand and Australian researchers. Additionally, he is the founding Co-Director of the Cyber Security Foundry, the first multi-disciplinary center in New Zealand for Cyber Security, focused on strengthening collaboration between industry and academia. From 2013 to 2014, Giovanni held the position of founding CEO at a startup targeting the smartphone security market.
His research interests include human-centered cyber security, policy-based security systems, privacy and confidentiality in cloud computing, smartphone security, and applied cryptography.
Talk Title: Human Centred Cyber Security
Phishing attacks are projected to incur a staggering cost of US$30 billion to the global economy in 2023 alone, with estimates indicating a continued rise. Over the next decade, expenses linked to ransomware—often resulting from successful phishing attacks—are anticipated to increase tenfold, reaching an annual total of US$300 billion.
Most of the current efforts to thwart phishing have concentrated on technological solutions or what we refer to as "blame-the-user" strategies. However, both approaches fall short. While technological measures have been effective, blocking approximately 90 percent of malicious emails, the remaining 10 percent—given the enormous volume of phishing emails (160 million per day)—still pose a significant threat. Likewise, user-based interventions, despite awareness initiatives aimed at training users to recognize suspicious emails, have not fully addressed the issue. Surprisingly, 65 percent of companies falling victim to phishing attacks had undergone some form of prior training.
In this talk, I will present some of the approaches developed with my collaborators, which emphasize the need to focus on individuals and the contexts in which they encounter and respond to phishing attacks.
Yong-Duan Song
Professor Yong-Duan Song is a Fellow of IEEE, Fellow of AAIA, Fellow of International Eurasian Academy of Sciences, and Fellow of Chinese Automation Association. He was one of the six Langley Distinguished Professors at National Institute of Aerospace (NIA), USA and register professional engineer (USA). He is currently the dean of Research Institute of Artificial Intelligence at Chongqing University. Professor Song is the Editor-in-Chief of IEEE Transactions on Neural Networks and Learning Systems (TNNLS) and the founding Editor-in-Chief of the International Journal of Automation and Intelligence.
Talk Title: Intermittent Sensoring and Control for Energy, Communication, and Computation Savings: Recent Developments and Future Trends
Utilizing discontinuous or intermittent feedback signals to generate intermittent control actions for nonlinear dynamic systems is both an intriguing and challenging topic. This presentation will provide an overview of recent advancements in various methods for intermittent sensing and control, with a particular focus on dynamic event-triggering techniques. Additionally, some of the latest findings from the speaker’s research group will be discussed.