top of page
Cheng Chin
Professor of Intelligent Systems Modelling & Simulation, Director of Research, Director of Newcastle University_Nvidia Joint Laboratory, Singapore
Bio: He is a Chair Professor of Intelligent Systems Modelling and Simulation at Newcastle University and Editor-in-Chief of Cybernetics and Systems. Since 2010, he has contributed to the Marine Engineering programme, focusing on marine electrical engineering, system modelling, and simulation. He is also an Adjunct Professor at Chongqing University, an NVIDIA Certified Instructor, and a Deep Learning Institute Ambassador. He leads the Newcastle University–NVIDIA Joint Laboratory and serves as Director of Research and Innovation in Singapore, where he has strengthened academia–industry collaboration. He has secured 9 EDB Industrial Postgraduate Programme grants and 2 Singapore Maritime Institute grants as PI/Co-PI, advancing intelligent system design and predictive analytics research.
He has published 5 books, edited 2, and holds multiple patents in Singapore and the US. His research in AI has been presented at major conferences, and he has received awards including Best Presentation at an IEEE conference, Best Paper at MIC, and the Judges’ Award at DCASE2019. He is a Fellow of the Higher Education Academy and has served on IEEE working groups contributing to IEEE 2660.1-2020 and IEEE 45.1-2023. He also serves on editorial boards of several international journals.
Title： Intelligent Systems Modelling
Abstract： Intelligent systems modelling has emerged as a powerful paradigm for representing, analysing, and controlling complex engineering infrastructures through the integration of data-driven learning, physical principles, and adaptive decision-making. In the context of water distribution networks (WDNs), such modelling frameworks enable enhanced situational awareness and predictive capabilities for efficient resource management. However, persistent leakages in WDNs lead to substantial water losses, underscoring the need for reliable and robust detection and localization mechanisms.
This presentation reviews existing leakage detection technologies from an intelligent systems perspective, encompassing model-based, data-driven, and hybrid methodologies. The strengths and limitations of current approaches are critically analysed, particularly in their ability to operate under real-world conditions. Key uncertainties such as sensor noise, demand fluctuations, and modelling inaccuracies, are systematically examined, as they significantly impact detection performance. To address these challenges, the presentation discusses advanced intelligent modelling strategies that incorporate adaptive learning, uncertainty quantification, and robust estimation techniques, aiming to improve the accuracy, resilience, and scalability of leakage detection and localization in WDNs.
Lei Chen
Shandong University, China
Bio: Lei Chen received the B.Sc. and M.Sc. degrees in electrical engineering from Shandong University, Jinan, China, and the Ph.D. degree in electrical and computer engineering from University of Ottawa, Ontario, Canada. He is currently an Associate Professor with the School of Information Science and Engineering, Shandong University, China. His research interests include image processing and computer vision, visual quality assessment and pattern recognition, machine learning and artificial intelligence. He was the principal investigator of projects granted from the National Natural Science Foundation of China, National Natural Science Foundation of Shandong Province, China Postdoctoral Science Foundation, etc. He has published more than 40 papers on top international journals and conferences in recent years including IEEE TIP, Signal Process., ICME, etc. He was awarded the Future Plan for Young Scholars of Shandong University. He served for many international conferences including the ICIGP 2021, CSAI2022, MLCCIM2022, and ICIVC 2023 as Program Chair, Technical Chair or Publicity Chair.
Title： Multi-Modal Spatio-Temporal Modeling Methods for Video Anomaly Detection
Abstract： In recent years, the video surveillance systems are widely used in the fields of urban safety, security management, crime-fighting, and healthcare. The research on abnormal behavior detection in video is crucial to maintain safety and improve the quality of life. However, surveillance environments often present severe conditions, such as fluctuating lighting, the presence of shadows, and adverse weather conditions. These background variations introduce noises for human behavior features and degrades the abnormal behavior detection performance. To address these problems, we propose a new framework called efficient abnormal behavior detection that simultaneously integrates spatio-temporal feature modeling and long-term dependency modeling. And we propose a cross-scale gated embedding graph model for skeleton-based anomaly detection to address the challenges of fusing cross-scale features in skeleton-based video data. The experimental results show the effectiveness of our proposed methods and demonstrate superiority over other related methods. The research findings can be used to identify and intervene in potential threats, accidents, and dangerous situations.
Philippe Fournier-Viger
Shenzhen University, China
Bio: Philippe Fournier-Viger (Ph.D) is distinguished professor at Shenzhen University (China). Five years after completing his Ph.D., he came to China in 2015 and became full professor after receiving a national talent title from the government of China. He has published more than 400 research papers related to data mining algorithms for complex data (sequences, graphs), intelligent systems and applications, which have received more than 18,500 citations (H-Index 66 - Google Scholar). He is the founder of the popular SPMF data mining library, offering more than 300 algorithms to find patterns in data, cited in more than 1,100 research papers. He is former associate editor-in-chief of the Applied Intelligence journal and has been keynote speaker for over 60 international conferences and co-edited four books for Springer. He is a book series editor for Atlantis Press (Springer). He appears in the top 0.3% of researchers for scientific influence in the Stanford list. He is an Elsevier “Highly Cited Chinese Researcher” since 2022. He won the "Most Influential Paper Award" at the 2024 PAKDD conference and received seven Best Paper Awards at international conferences. He has been general chair or conference chair for 10 international conferences, program committee chair for 15 conferences, and organized more than 25 workshops and special sessions at international conferences. He has collaborated with over 400 researchers worldwide. Website: http://www.philippe-fournier-viger.com.
Title： Algorithms for Identifying Interesting Structures in Complex Data
Abstract： Industry and society are changing rapidly with progress in artificial intelligence, driven by advances in machine learning, cloud computing, large-scale data storage, distributed systems, and modern hardware such as GPUs and TPUs. These technologies enable the collection and processing of massive amounts of data, making data a central element in intelligent systems. However, beyond raw data, a key challenge is that data often contains hidden structures and patterns that are not directly visible but can reveal important insights across different domains. For example, in social networks, they can describe interaction behaviors and communities; in software API call data, they can indicate abnormal or malicious activities; and in trajectory data, they can capture behavioral regularities, movement patterns, recurring trajectories, as well as anomalies and outliers. This raises a fundamental question: how can we effectively analyze large-scale data to uncover interesting hidden structures? This talk addresses this problem through data mining and pattern mining algorithms, which aim to extract meaningful and interpretable patterns from complex datasets. We begin by describing the overall process of pattern discovery, and then describe recent applications in bioinformatics, malware detection, network alarm analysis, along with the SPMF open-source software as a practical tool for pattern discovery. Finally, we highlight how pattern mining can complement modern AI methods to better understand complex systems and support future intelligent applications.
Kannimuthu Subramaniyam
Karpagam College of Engineering, India
Bio: Kannimuthu Subramaniyam is currently working as Professor in the Department of Computer Science and Engineering at Karpagam College of Engineering, Coimbatore, Tamil Nadu, India. He is also an In-Charge for the Center of Excellence in Algorithms. He is an IBM Certified Cybersecurity Analyst. He did PhD in Computer Science and Engineering at Anna University, Chennai. He did his M.E (CSE) and B.Tech (IT) at Anna University, Chennai. He has more than 16 years of teaching and industrial experience. He is the recognized supervisor of Anna University, Chennai. Three PhD candidate is completed their research under his guidance. He is now guiding 11 PhD Research Scholars. He has published 59 research articles in various International Journals. He published 2 books ("Artificial Intelligence" & “LinkedList Demystified-A Placement Perspective” and 3 Book Chapters (WOS / Scopus Indexed). He is acting as mentor / consultant for DeepLearning.AI, Hubino, MaxByte Technologies and Dhanvi Info Tech, Coimbatore. He is the expert member for AICTE Student learning Assessment Project (ASLAP). He has presented a number of papers in various National and International conferences. He has visited more than 100 Engineering colleges and delivered more than 138 Guest Lectures on various topics. He is the reviewer for 50 Journals and 3 Books. He has successfully completed the consultancy project through Industry-Institute Interaction for ZF Wind Power Antwerpen Ltd., Belgium. He has received funds from CSIR, DRDO and ISRO to conduct workshops and seminars. He has completed more than 610 Certifications (41 Specializations and 4 Professional Certifications) in Coursera, Hackerrank and NPTEL on various domains. He has guided a number of research-oriented as well as application-oriented projects organized by well-known companies like IBM. He is actively involving in setting up lab for Cloud Computing, Big Data Analytics, Open-Source Software, Internet Technologies etc., His research interests include Artificial Intelligence, Data Structures and Algorithms, Machine Learning, Big Data Analytics, Virtual Reality & Blockchain. One of his research works is incorporated SPMF Open-Source Data Mining Tool. Source: http://www.philippe-fournier-viger.com/spmf/index.php?link=algorithms.php. He Conferred Second Best Team in NLP Challenge as part of FIRE 2019 conference. He secured first Position in NLP Challenge as part of FIRE 2018 Conference.
Title： Certain Investigations on Hybrid Deep Learning Algorithms for Autism Spectrum Disorder Diagnosis
Abstract： Autism Spectrum Disorder (ASD) is a complex neurodevelopmental condition that affects communication, social interaction, and behavioural patterns. Individuals with ASD often exhibit repetitive actions, atypical facial expressions, and difficulties in social engagement. While therapies can help manage symptoms, ASD remains incurable, and its severity often leads to delayed or missed diagnoses. Accurate and early prediction is therefore critical, yet conventional diagnostic methods—such as behavioral assessments, questionnaires, and manual evaluations—are often subjective, time-consuming, and prone to inconsistencies.
Recent advances in artificial intelligence have opened new avenues for objective and efficient ASD diagnosis by leveraging multimodal data sources such as fMRI scans, eye-tracking, motor activity, and behavioral kinematics. In this keynote, I present our exploration of five hybrid deep learning models—MobileNetV2+BiLSTM, ResNet50+LSTM, EfficientNetB4, InceptionV3, and MobileNetV2+GRU—evaluated on image data from the Kaggle repository. Among these, the MobileNetV2+GRU model demonstrated superior predictive performance.
The hybrid framework combines the lightweight MobileNetV2 convolutional neural network, responsible for extracting salient visual features with minimal computational cost, and a Gated Recurrent Unit (GRU) layer, which effectively captures temporal dependencies in the data. This synergy enables robust feature learning while maintaining efficiency. Comparative analysis with other architectures highlights the MobileNetV2+GRU model’s ability to outperform alternatives, achieving a test accuracy of 95.66%.
These findings underscore the potential of hybrid deep learning approaches to enhance the reliability, efficiency, and accessibility of ASD diagnosis, particularly in children. The keynote will further discuss the implications of integrating such AI-driven diagnostic tools into clinical practice, paving the way for earlier interventions and improved quality of life for individuals with ASD.
bottom of page