Keynote Speakers

 

       Prof. Robert Minasian, IEEE & OSA Fellow, The University of Sydney, Australia

 

 

Professor Minasian is a Chair Professor with the School of Electrical and Information Engineering at the University of Sydney, Australia. He is also the Director of the Fibre-optics and Photonics Laboratory. His research has made key contributions to microwave photonic signal processing. He is recognized as an author of one of the top 1% most highly cited papers in his field worldwide. Professor Minasian has contributed over 360 research publications, including Invited papers in the IEEE Transactions and Journals, and Plenary and Invited papers at leading international conferences. Professor Minasian was the recipient of the ATERB Medal for Outstanding Investigator in Telecommunications, awarded by the Australian Telecommunications and Electronics Research Board. He is a Fellow of the IEEE, and a Fellow of the Optical Society of America.

Speech Title: Progress in photonic signal processing

Abstract: Next generation signal processing systems will require new technologies to address the current limitations for growth in capacity and versatility in areas such as 5G and Internet of Things (IoT). Microwave photonics, which merges the worlds of RF and photonics, shows strong potential as a key enabling technology to enable new paradigms in the processing of high speed signals and in sensing that can overcome inherent electronic limitations. Photonic signal processors are intrinsically compatible with optical-wireless systems, and can provide connectivity with in-built signal conditioning, while also providing important advantages of EMI immunity. Moreover, photonic integration on semiconductor material platforms that co-exist with CMOS electronics enables boosting the performance of future systems performing sensing and communications with the potential for implementing high bandwidth, fast and complex functionalities. Recent photonic signal processing advances will be presented including widely tunable single passband filters, high-speed frequency converters, and a range of novel high-resolution sensors based on microwave photonic techniques.

 

       Prof. Mao Kezhi, Nanyang Technological University, Singapore

 

 

Dr. Mao Kezhi obtained his BEng, MEng and PhD from Jinan University, Northeastern University, and University of Sheffield in 1989, 1992 and 1998 respectively. He joined School of Electrical and Electronic Engineering, Nanyang Technological University, Singapore in 1998, where he is now a tenured Associate Professor. Dr. Mao has over 20 years of research experience in artificial intelligence, machine learning, big data, image processing, natural language processing, and information fusion. He has published over 100 research papers in referred international journals and conferences. He has edited 3 books published by Springer.
Besides academic research, Dr. Mao is also active in development and consulting. He has successfully developed and delivered several intelligent systems and software tools to government agencies, hospitals and industries.
Dr. Mao Kezhi serves on Editorial Board of Computational Intelligence and Neuroscience. He has served as Keynote Speaker/Programme Chair/Organizing Committee Member for multiple international conferences. In addition, he has served as a reviewer for multiple international journals.

Speech Title: Knowledge-oriented convolutional neural network (K-CNN) and its application in causal relation extraction

Abstract: The cause-effect relation plays an important role in human cognition due to its significant impact on reasoning and decision making. However, causal relation extraction is a very challenging task in Natural Language Processing (NLP). There are many existing approaches developed to tackle this task, either rule-based (non-statistical) or machine-learning-based (statistical) method. This talk will present a knowledge-oriented convolutional neural network (K-CNN) approach that integrates knowledge and data. The proposed K-CNN consists of two channels including knowledge-oriented channel and data-oriented channel. The knowledge-oriented channel incorporates human knowledge to capture the linguistic clues of causal relationship, while the data-oriented channel leans important features of causal relation from training data. The convolutional filters in knowledge-oriented channel are automatically generated from open knowledge bases such as WordNet and FrameNet without involving training. In addition, two types of latent semantic features including WordNet categorical features and FrameNet causal scores are proposed to allow the model to capture more useful latent information that cannot be captured from data by CNN itself.

 

       Prof. Xudong Jiang, Nanyang Technological University, Singapore

 

 

Prof. Xudong Jiang received the B.Sc. and M.Sc. degree from the University of Electronic Science and Technology of China, in 1983 and 1986, respectively, and received the Ph.D. degree from Helmut Schmidt University Hamburg, Germany in 1997, all in electrical and electronic engineering. From 1986 to 1993, he worked as Lecturer at the University of Electronic Science and Technology of China where he received two Science and Technology Awards from the Ministry for Electronic Industry of China. He was a recipient of the German Konrad-Adenauer Foundation young scientist scholarship. From 1993 to 1997, he was with Helmut Schmidt University Hamburg, Germany as scientific assistant. From 1998 to 2004, He worked with the Institute for Infocomm Research, A*Star, Singapore, as Senior Research Fellow, Lead Scientist and appointed as the Head of Biometrics Laboratory where he developed a fingerprint verification algorithm that achieved the fastest and the second most accurate fingerprint verification in the International Fingerprint Verification Competition (FVC2000). He joined Nanyang Technological University, Singapore as a faculty member in 2004 and served as the Director of the Centre for Information Security from 2005 to 2011. Currently, Dr Jiang is a tenured Associate Professor in School of Electrical and Electronic Engineering, Nanyang Technological University. Dr Jiang has published over hundred research papers in international refereed journals and conferences, some of which are well cited on Web of Science. He is also an inventor of 7 patents (3 US patents), some of which were commercialized. Dr Jiang is a senior member of IEEE and has been serving as Editorial Board Member,Guest Editor and Reviewer of multiple international journals, and serving as Program Committee Chair, Keynote Speaker and Session Chair of multiple international conferences. His research interest includes pattern recognition, computer vision, machine learning, image analysis, signal/image processing, machine learning and biometrics.

Speech Title: The Role of Dimensionality Reduction for Classification

Abstract: Finding/extracting low-dimensional structures in high-dimensional data is of increasing importance, where data/signals lie in observational spaces of thousands, millions or billions of dimensions. Many biomedical and bioinformatics datasets have very high dimension and low number of samples. The curse of dimensionality is in full play here: We have to conduct inference with a limited or no human knowledge. Machine learning is a solution that becomes hotter and hotter to boiling. This is evidenced by numerous techniques published in the past decade, many of which are in prestige journals. Nevertheless, there are some fundamental concepts and issues still unclear or in paradox. For example, we often need many processing steps in a complex information discovery/recognition system. As the information amount cannot be increased and must be reduced by any processing, why do we need it before the main processing? This seemly simple question easily answerable if each step uses different prior knowledge is nontrivial in machine learning. People proposed numerous machine learning approaches but seem either unaware of or avoiding this fundamental issue. Although extracting the most discriminative information is indisputably the ultimate objective for pattern recognition, this talk will challenge it as a proper or effective criterion for the machine learning-based dimension reduction or information/feature extraction, despite the fact that it has been employed by almost all researchers.