BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Department of Electrical and Computer Engineering (HKUECE) 電機與計算機工程系 - ECPv6.15.20//NONSGML v1.0//EN
CALSCALE:GREGORIAN
METHOD:PUBLISH
X-WR-CALNAME:Department of Electrical and Computer Engineering (HKUECE) 電機與計算機工程系
X-ORIGINAL-URL:https://ece.hku.hk
X-WR-CALDESC:Events for Department of Electrical and Computer Engineering (HKUECE) 電機與計算機工程系
REFRESH-INTERVAL;VALUE=DURATION:PT1H
X-Robots-Tag:noindex
X-PUBLISHED-TTL:PT1H
BEGIN:VTIMEZONE
TZID:Asia/Hong_Kong
BEGIN:STANDARD
TZOFFSETFROM:+0800
TZOFFSETTO:+0800
TZNAME:HKT
DTSTART:20230101T000000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTART;TZID=Asia/Hong_Kong:20241205T110000
DTEND;TZID=Asia/Hong_Kong:20241205T120000
DTSTAMP:20260510T152019
CREATED:20241126T020836Z
LAST-MODIFIED:20250211T042315Z
UID:19487-1733396400-1733400000@ece.hku.hk
SUMMARY:Empowering High-Performance Computing Through Heterogeneous Integration of Power Electronics
DESCRIPTION:Abstract\nMost power electronic equipment\, including power supplies for Artificial Intelligence (AI) and high-performance computing systems\, has traditionally been designed using discrete active and passive components. However\, the power electronics industry has reached a stage where improving one performance attribute often comes at the expense of others. As AI and high-performance computing systems demand increasingly higher power levels\, existing power delivery architectures are proving inadequate. Delivering the required power to server racks\, circuit boards\, accelerator cards\, and AI processors has become an escalating challenge. The emergence of wide-bandgap (WBG) power semiconductor devices\, such as silicon carbide (SiC) and gallium nitride (GaN)\, presents a breakthrough opportunity. Compared to silicon (Si) devices\, WBG technologies offer significantly lower losses\, providing a promising solution to the energy challenges of high-performance computing. However\, current design practices often adopt a ‘plug-and-play’ approach\, merely replacing Si with WBG components without altering the underlying design principles. This results in only incremental improvements in efficiency and power density\, leaving the transformative potential of WBG devices untapped. This presentation will explore a paradigm shift in powering AI and high-performance computing systems through the use of high-frequency heterogeneous integration in WBG power electronics design. This innovative approach delivers simultaneous advancements in all critical performance metrics\, including efficiency\, power density\, and electromagnetic compatibility (EMC). Furthermore\, this approach can streamline traditionally labor-intensive manufacturing processes\, paving the way for significant advancements in overall production efficiency. \nSpeaker\nProf. Qiang Li\nFull Professor\,\nCenter for Power Electronics Systems (CPES)\,\nVirginia Tech \nBiography of the Speaker\nQiang Li received the B.S. and M.S. degrees from Zhejiang University\, China\, in 2003 and 2006\, respectively\, and the Ph.D. degree from Virginia Tech\, Blacksburg\, VA\, in 2011. He is currently a Full Professor in the Center for Power Electronics Systems (CPES) at Virginia Tech. His research interests include high-frequency power conversion and control\, high-density electronics packaging and magnetic integration\, as well as power solutions for high-performance computing\, data centers\, electric vehicles\, and energy storage systems. With over 300 peer-reviewed technical publications\, including 100 journal articles\, he has received eight prize paper awards and holds 26 U.S. patents. He currently serves as the Chair of Academic Affairs for the IEEE Power Electronics Society and is an associate editor for both the IEEE Transactions on Power Electronics and the IEEE Journal of Emerging and Selected Topics in Power Electronics. Dr. Li is also a recipient of the U.S. National Science Foundation (NSF) Career Award. \nOrganiser\nProf. Han Wang \nAll are welcome!
URL:https://ece.hku.hk/events/20241205-2/
LOCATION:Room CB-603\, 6/F\, Chow Yei Ching Building\, The University of Hong Kong
CATEGORIES:Highlights,Seminar
ATTACH;FMTTYPE=image/jpeg:https://ece.hku.hk/wp-content/uploads/2024/11/1280-copy.jpg
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Hong_Kong:20241205T150000
DTEND;TZID=Asia/Hong_Kong:20241205T160000
DTSTAMP:20260510T152019
CREATED:20241120T071853Z
LAST-MODIFIED:20250211T042315Z
UID:19462-1733410800-1733414400@ece.hku.hk
SUMMARY:Multimode Fiber Communication and Multicore Fiber Endoscopy Exploiting Physics-informed Deep Neural Networks
DESCRIPTION:Abstract\nAdvances in fiber-optic systems are crucial for the internet. To cope with this continued exponential growth as Keck´s law\, multimode fibers (MMF) with spatial division multiplexing (SDM) are proposed. However\, there are challenging mode scattering effects in MMF. Physics-informed deep learning enables to correct the scattering\, resulting in advancing security and data rate. We highlight also 3D imaging with lensless multicore fiber endoscopes exploiting learnable multiple Wiener net. Multimode fiber communication and multicore fiber endoscopy are promising for advancements of the internet of things and of biomedical diagnostics and therapy. \nSpeaker\nProf. Jürgen Czarske\nDirector and Full Chair Professor，\nTU Dresden\, Germany \nBiography of the Speaker\nJuergen W Czarske (Fellow EOS\, OPTICA\, SPIE\, IET\, IOP) is director and full chair professor of the TU Dresden\, Germany. His awards include the 2019 OPTICA Joseph-Fraunhofer-Award/Robert-M.-Burley-Prize in Optical Engineering and the 2024 SPIE Dennis Gabor Award in Diffractive Optics. Juergen fosters talented students early. The students and members of his lab have won over 100 prizes\, including Bertha-Benz award of Daimler Benz Foundation (10\,000 Euro). Juergen is Vice President of International Commission for Optics\, ICO\, and was the general chair of the world congress ICO-25\, which was co-sponsored by OPTICA\, SPIE\, IEEE\, Zeiss\, DGaO-The German Branch of EOS\, IUPAP. 3 Nobel laureates have delivered plenary lectures and the participants came from 5A (America\, Asia\, Australia\, Africa and Amazing Europe). \nOrganiser\nProf. Kevin K.M. Tsia\nDepartment of Electrical and Electronic Engineering\,\nThe University of Hong Kong \nSupported by\nInnovation Wing Two \nAll are welcome!
URL:https://ece.hku.hk/events/20241205-1/
LOCATION:Tam Wing Fan Innovation Wing Two\, G/F\, Run Run Shaw Building\, The University of Hong Kong
CATEGORIES:Highlights,Seminar
ATTACH;FMTTYPE=image/jpeg:https://ece.hku.hk/wp-content/uploads/2025/01/1280-2.jpg
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Hong_Kong:20241209T140000
DTEND;TZID=Asia/Hong_Kong:20241209T150000
DTSTAMP:20260510T152019
CREATED:20241205T041102Z
LAST-MODIFIED:20250211T042315Z
UID:19528-1733752800-1733756400@ece.hku.hk
SUMMARY:Advancing Accessible Ultra-low-field MRI with Deep Learning
DESCRIPTION:Abstract\nMagnetic Resonance Imaging (MRI) is a versatile medical imaging modality. Despite its crucial role in modern healthcare\, MRI remains largely inaccessible to the general population. The development of ultra-low-field (ULF) MRI offers opportunities for enabling low-cost\, low-power\, and potentially portable clinical applications. However\, the imaging performance of these emerging ULF MRI scanners remains poor due to the significantly lower signal-to-noise ratio resulting from the weaker magnetic field. Recent advancements in deep learning have opened new frontiers to tackle this unique challenge. This talk will introduce a paradigm shift in improving ULF MR image quality and reconstructing MR image through supervised deep learning techniques. Leveraging existing high-quality high-field MRI data for training\, DL models can recover structural details buried in noise\, artefacts and poor resolution of raw ULF MR images. Such approaches have the potential to overcome the limitations of poor image quality and pave the way for clinical adoption of ULF MRI. \nSpeaker\nDr. Vick Lau\nDepartment of Electrical and Electronic Engineering\,\nThe University of Hong Kong \nBiography of the Speaker\nDr. Vick Lau obtained his MEng degree in Biomedical Engineering from Imperial College London in 2019. He recently received his PhD degree in Electrical and Electronic Engineering from the University of Hong Kong in 2024. His research focuses on the application of deep learning techniques to MRI image processing\, restoration\, reconstruction and analysis\, particularly for ULF and accessible MRI. \nOrganiser\nProf. Ed X. Wu\nDepartment of Electrical and Electronic Engineering\,\nThe University of Hong Kong \nAll are welcome!
URL:https://ece.hku.hk/events/20241209-1/
LOCATION:Room CB-603\, 6/F\, Chow Yei Ching Building\, The University of Hong Kong
CATEGORIES:Highlights,Seminar
ATTACH;FMTTYPE=image/jpeg:https://ece.hku.hk/wp-content/uploads/2024/12/1280.jpg
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Hong_Kong:20241218T103000
DTEND;TZID=Asia/Hong_Kong:20241218T113000
DTSTAMP:20260510T152019
CREATED:20241210T012223Z
LAST-MODIFIED:20250211T042315Z
UID:107413-1734517800-1734521400@ece.hku.hk
SUMMARY:Unlocking the Value of Single Modality through Multi-Modal Knowledge Transfer for Healthcare
DESCRIPTION:Abstract\nRecent years have witnessed the remarkable success of deep neural networks in healthcare\, particularly in the analysis of medical images and signals. However\, their performance is often constrained by the scarcity of labelled data\, driven by high labelling costs and challenges related to data privacy and sharing. In this talk\, I will explore how we can overcome these limitations by leveraging multi-modal data through advanced learning frameworks to enhance the capabilities of single modality analysis. Specifically\, I will present our recent and ongoing work\, including those accepted at MICCAI 2023 and MICCAI 2024. This talk will delve into the details of innovative techniques such as large language model-informed pretraining and multi-modal learning for X-ray images and ECG signals\, as well as demonstrating how these approaches can significantly contribute to more accurate\, reliable\, and cost-effective healthcare solutions. \nSpeaker\nProf. Chen (Cherise) Chen\nLecturer (Assistant Professor) in Computer Vision\,\nSchool of Computer Science\,\nUniversity of Sheffield\, UK \nBiography of the Speaker\nChen (Cherise) Chen is currently a Lecturer (assistant professor) in Computer Vision at the School of Computer Science\, University of Sheffield\, UK. Previously\, she was a postdoc at Imperial College London (ICL) and then the University of Oxford. She obtained her MSc and Ph.D. from the Department of Computing at Imperial College London in 2016 and 2022\, respectively\, where she worked closely with Prof. Daniel Rueckert and Dr. Wenjia Bai. Chen also has accumulated valuable industrial experience. She worked as a research scientist at Infervision Inc. in Beijing in 2017\, prior to her PhD\, and later as a part-time research scientist at HeartFlow\, UK\, in 2022 following her PhD. Her research focuses on the intersection of AI and healthcare\, particularly in developing data-efficient\, robust\, and explainable AI for clinical applications. So far\, she has published more than 40 papers in leading conferences and high-impact journals on deep learning for medical data analysis such as MICCAI\, ECCV\, IEEE TMI\, and Medical Image Analysis\, accumulating over 2\,000 Google Scholar citations and an h-index of 20. She is a program chair for MIDL 2025; session and area chair for MICCAI 2024 and serves as lead organisers in several MICCAI workshops and challenges including\, MICCAI ADSMI 2024\, DALI 2023\, and the CMRxMotion Challenge. Very recently\, she has also been appointed as an ELLIS Scholar at the European Laboratory for Learning and Intelligent Systems in 2024. https://cherise215.github.io. \nOrganiser\nProf. Cheng Chen\nDepartment of Electrical and Electronic Engineering\,\nThe University of Hong Kong \nAll are welcome!
URL:https://ece.hku.hk/events/20241218-2/
LOCATION:Room CB-603\, 6/F\, Chow Yei Ching Building\, The University of Hong Kong
CATEGORIES:Highlights,Seminar
ATTACH;FMTTYPE=image/jpeg:https://ece.hku.hk/wp-content/uploads/2024/12/1280-1.jpg
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Hong_Kong:20241218T150000
DTEND;TZID=Asia/Hong_Kong:20241218T160000
DTSTAMP:20260510T152019
CREATED:20241128T022442Z
LAST-MODIFIED:20250211T042315Z
UID:19513-1734534000-1734537600@ece.hku.hk
SUMMARY:Toward Scalable Generative AI via Mixture of Experts in Mobile Edge Networks
DESCRIPTION:Abstract\nThe evolution of generative artificial intelligence (GAI) has driven revolutionary applications like ChatGPT. The proliferation of these applications is underpinned by the mixture of experts (MoE)\, which contains multiple experts and selectively engages them for each task to lower operation costs while maintaining performance. Despite MoE’s efficiencies\, GAI still faces challenges in resource utilization when deployed on local user devices. Therefore\, we first propose mobile edge networks supported MoE-based GAI. Rigorously\, we review the MoE from traditional AI and GAI perspectives\, scrutinizing its structure\, principles\, and applications. Next\, we present a new framework for using MoE for GAI services in Metaverse. Moreover\, we propose a framework that transfers subtasks to devices in mobile edge networks\, aiding GAI model operation on user devices. Moreover\, we introduce a novel approach utilizing MoE\, augmented with Large Language Models (LLMs)\, to analyze user objectives and constraints of optimization problems based on deep reinforcement learning (DRL) effectively. This approach selects specialized DRL experts\, and weights each decision from the participating experts. In this process\, the LLM acts as the gate network to oversee the expert models\, facilitating a collective of experts to tackle a wide range of new tasks. Furthermore\, it can also leverage LLM’s advanced reasoning capabilities to manage the output of experts for joint decisions. Lastly\, we insightfully identify research opportunities of MoE and mobile edge networks. \nSpeaker\nProf. Dusit Niyato\nPresident’s Chair Professor\,\nCollege of Computing & Data Science (CCDS)\,\nNanyang Technological University\, Singapore \nBiography of the Speaker\nDusit Niyato is a President’s Chair Professor in the College of Computing & Data Science (CCDS)\, Nanyang Technological University\, Singapore. Dusit’s research interests are in the areas of mobile generative AI\, edge intelligence\, quantum computing and networking\, and incentive mechanism design. Dusit won the IEEE Vehicular Technology Society Stuart Meyer Memorial Award. Dusit won the IEEE Vehicular Technology Society Stuart Meyer Memorial Award. Currently\, Dusit is serving as Editor-in-Chief of IEEE Communications Surveys and Tutorials (impact factor of 34.4 for 2023) and will serve as the Editor-in-Chief of IEEE Transactions on Network Science and Engineering (TNSE) from 2025. He is also an area editor of IEEE Transactions on Vehicular Technology (TVT)\, topical editor of IEEE Internet of Things Journal (IoTJ)\, lead series editor of IEEE Communications Magazine\, and associate editor of IEEE Transactions on Wireless Communications (TWC)\, IEEE Transactions on Mobile Computing (TMC)\, IEEE Wireless Communications\, IEEE Network\, IEEE Transactions on Information Forensics and Security (TIFS)\, IEEE Transactions on Cognitive Communications and Networking (TCCN)\, IEEE Transactions on Services Computing (TSC)\, and ACM Computing Surveys. Dusit is the Members-at-Large to the Board of Governors of IEEE Communications Society for 2024-2026. He was named the 2017-2023 highly cited researcher in computer science. He is a Fellow of IEEE and a Fellow of IET. \nOrganiser\nProf. Hongyang Du\nDepartment of Electrical and Electronic Engineering\,\nThe University of Hong Kong \nAll are welcome!
URL:https://ece.hku.hk/events/20241218-1/
LOCATION:Room CB-603\, 6/F\, Chow Yei Ching Building\, The University of Hong Kong
CATEGORIES:Highlights,Seminar
ATTACH;FMTTYPE=image/jpeg:https://ece.hku.hk/wp-content/uploads/2024/11/1280.jpg
END:VEVENT
END:VCALENDAR