Yu-Chee Tseng received his Ph.D. in Computer and Information Science from the Ohio State University in January of 1994. He served as Chairman (2005-2009) and as Dean (2011-2017) for College of Computer Science, National Yang Ming Chiao Tung University (NYCU), Taiwan. He is founding Dean of College of AI, NYCU. Currently, he is Director of Pervasive AI Research Labs, NYCU. Dr. Tseng has been awarded as NYCU Chair Professor (2011-present) and Y. Z. Hsu Scientific Chair Professor (2012-2013). He received Outstanding Research Award (National Science Council, 2001, 2003, and 2009), Academic Award (Ministry of Education), Elite I. T. Award (2004), and Distinguished Alumnus Award (Ohio State University, 2005), Y. Z. Hsu Scientific Paper Award (2009), TWAS Prize (2018), and National Chair Professorship (2020-2023). His research interests include mobile computing, wireless communication, and AI. Dr. Tseng is an IEEE Fellow. He served on the editorial boards of IEEE Trans. on Vehicular Technology, IEEE Trans. on Mobile Computing, IEEE Trans. on Parallel and Distributed Systems, and IEEE Internet of Things Journal. His citation h-index is 68.

Title: Exploring Modalities in Machine Learning
Abstract:A multi-modal deep learning model can process and interpret multiple types of data, such as text, images, audio, and video. These models are significantly closer to human cognition capability when perceiving the world. Let A, B, and C be three modalities. They may interact with each other during model training. Recent researches have made success in several types of modality interactions: (1) alignment of modality A and modality B, denoted as A~B, (2) conversion from modality A to modality B, denoted as A→B, (3) merging a model of A and a model of B into a larger model C, denoted as AUB=C, (4) inject a new modality C into two existing modalities A and B, denoted as A+B→C, (5) transferring the calculus on modality A to modality B, denoted as A→(+, -, *, /)→B, (6) producing a secondary modality A’ from A, denoted as A→A’, and (7) stacking two modalities A and B, denoted as A//B. We will illustrate several examples for the above interaction types. Understanding these may help develop new researches.

Ai-Chun Pang (逄愛君) received B.S., M.S., and Ph.D. degrees in Computer Science and Information Engineering from National Chiao Tung University (now National Yang Ming Chiao Tung University), Taiwan, in 1996, 1998, and 2002, respectively. Dr. Pang is the Director and Distinguished Research Fellow of the Research Center of Information Technology Innovation, Academia Sinica. She is also the Distinguished Professor of the Department of Computer Science and Information Engineering at National Taiwan University. Her research interests include wireless and mobile networking, edge computing, and IoT. She is a Fellow of the IEEE.

Title: Toward 6G-Enabled Mobile Edge Intelligence
Abstract:With the explosive development of AI, edge intelligence has been considered a must in developing future 6G mobile communications systems to provide timely responses to emerging applications on mobile devices. In 6G, the computation-intensive AI tasks will be distributed at the network edge, and the communications paradigm will shift from conventional symbol transmission to semantic information delivery. This lecture will overview key features of 6G mobile networks and elaborate on distributed AI learning driven by edge intelligence. We will present the effects of limited labeled and non-IID data in the edge-intelligence environment. We will also discuss the vulnerability of the edge-intelligence framework and defense methods against privacy leakage and security threats. Finally, we will introduce the GenAI-based semantic encoder to prioritize task-oriented communication, with the concept of understanding before transmitting and delivering the intended meaning of messages, to achieve the goal of pervasive computing for connected intelligence.

Professor Mohammad S. Obaidat is an internationally known academic/researcher/scientist/ scholar. He received his Ph.D. degree in Computer Engineering with a minor in Computer Science from The Ohio State University, Columbus, USA. He has received extensive research funding and published To Date (2023) over One Thousand and Two Hundred (1,200) refereed technical articles-About half of them are journal articles, over 100 books, and 70 Book Chapters. He is Editor-in-Chief of 3 scholarly journals and an editor of many other international journals. He is the founding Editor-in Chief of Wiley Security and Privacy Journal. Moreover, he is the founder or co-founder of 5 International Conferences. Among his previous positions are Advisor to the President of Philadelphia University for Research, Development and Information Technology, President and Chair of Board of Directors of the Society for Molding and Simulation International (SCS), Senior Vice President of SCS, SCS VP for Membership and SCS VP for Conferences, Dean of the College of Engineering at Prince Sultan University, Founding Dean of the College of Computing and Informatics at the University of Sharjah, Chair and tenured Professor at the Department of Computer and Information Science and Director of the MS Graduate Program in Data Analytics at Fordham University, Chair and tenured Professor of the Department of Computer Science and Director of the Graduate Program at Monmouth University, Chair and Professor of Computer Science Department at University of Texas-Permian Basin, Distinguished Professor at Indian Institute of Technology (IIT)-Dhanbad.

Wei-Chao Chen is the Chief Digital Officer at Inventec Corporation and the Chairman at Skywatch Innovation. His research interests include graphics hardware, computational photography, augmented reality, and computer vision. Dr. Chen was the Chief AI Advisor at Inventec (2018-2020), an adjunct faculty at the National Taiwan University (2009-2018), a senior research scientist in Nokia Research Center at Palo Alto (2007-2009), and a 3D Graphics Architect in NVIDIA (2002-2006). Dr. Chen received his MS in Electrical Engineering from National Taiwan University (1996) and Ph.D. in Computer Science from the University of North Carolina at Chapel Hill (2002).

Title: Reliable AI for Manufacturing: Challenges in Data Curation, Continual Learning, and Agile Production
Abstract:
Lights-out manufacturing refers to a production process requiring little human intervention. This promise of higher production efficiency also brings significant investment and poses multifaceted challenges for AI and robotics technologies. Data curation emerges as the primary obstacle, as the ability to collect, label, and evolve the datasets can significantly vary across various sites and parties. We also find it challenging to share data across parties because of the multiple stakeholders involved. Continual learning becomes critical as we encounter scope and concept changes during model deployment. Scaling out also challenges model reliability under domain transfer. Agility has become a key differentiator under dynamic production settings. The production facilities must react and adapt to new tasks for smaller production batches with greater versatility. As researchers and practitioners with firsthand experiences, we hope this talk can shed some light on these respective challenges and realize that we are still just a few steps away from foundation models in reliable AI manufacturing.

Wei-Bin Lee is the CEO of the Information Security Center at HonHai Research Institute. He has previously served as the Commissioner of the Department of Information Technology for the Taipei City Government, Chief Digital Officer at Taipei Fubon Bank, and Director of the Innovation and Technology Office at Fubon Financial Holding. Additionally, he has served as Chairman of the Artificial Intelligence Foundation. He has also been a Professor in the Department of Information Engineering and Computer Science at Feng Chia University. Dr. Lee's expertise lies in network security, cryptography, digital watermarking, and information security management. With his extensive experience, he is poised to lead HonHai Research Institute in making significant contributions to both industry and society. Dr. Lee holds a Ph.D. in Computer Science and Information Engineering from National Chung Cheng University, Taiwan.

Title: Digital Pandora: Reshaping Cybersecurity Landscape with Al and Quantum

Ee-Chien Chang is an Associate Professor in the School of Computing at National University of Singapore. He received his PhD in Computer Science from New York University, and was a postdoctoral fellow with DIMACS in Rutgers University and NEC Labs America. His research areas cover information security, multimedia, and their intersection. His earlier works include image forensic, image watermarking and secure cryptographic techniques for noisy data. More recently, he has been investigating issues in data privacy and cloud security. He has published in reputable conferences and journals, including CCS, EUROCRYPT, USENIX Security, ACM Multimedia, INFOCOM, Journal of Applied and Computational Harmonic Analysis, etc. He is a lead-PI of National Cybersecurity R&D Laboratory.

Title: Digital Watermarking into AI Era
Abstract:
Recent disruptive advancements in Machine Learning have found diverse applications and prompted new approaches in well-established research problems. One such area is digital watermarking, an active field that have been extensively studied over past two decades, starting from the pioneering work on spread spectrum watermarking in 1996. More recently, there has been significant research on watermarking with AI, which can be broadly categorized into two main categories: Using AI for watermarking and watermarking of AI. Many designs of watermarking encoders and decoders rely on intriguing mathematical coding structure in high dimensional space. Interestingly, Machine Learning can now be leveraged to enhance watermarking by “training” more effective encoders/decoders. This talk will provide an overview on Machine Learning approaches in enhancing watermarking performance, highlighting some of our contributions. Furthermore, watermarking of AI -- both AI-Generated Content (AIGC) and the AI models -- has become increasingly important, particularly for attribution, authentication, and intellectual property protection. While conventional watermarking techniques could be directly, the controllable nature of AI generation and the functionality of AI models offers new opportunities and challenges. We will also give a brief overview on such issues.