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-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:20250101T000000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTART;TZID=Asia/Hong_Kong:20260130T100000
DTEND;TZID=Asia/Hong_Kong:20260130T110000
DTSTAMP:20260511T172843
CREATED:20260126T040445Z
LAST-MODIFIED:20260126T040445Z
UID:114690-1769767200-1769770800@ece.hku.hk
SUMMARY:RPG Seminar – New Paradigm for Universal Graph Prompt Tuning
DESCRIPTION:Zoom Link:\nhttps://hku.zoom.us/j/92120929746?pwd=aL2f7KSR1uAb9rKfYOyaxpOKa9SBQL.1 \nAbstract\nWe strengthen the theoretical foundation of universal graph prompt tuning by introducing stricter constraints\, demonstrating that adding prompts to all nodes is a necessary condition for achieving the universality of graph prompts. To this end\, we propose a novel model and paradigm\, Learning and Editing Universal GrAph Prompt Tuning (LEAP)\, which preserves the theoretical foundation of universal graph prompt tuning while pursuing more ideal prompts. Specifically\, we first build the basic universal graph prompts to preserve the theoretical foundation and then employ actor-critic reinforcement learning to select nodes and edit prompts. Extensive experiments on graph- and node-level tasks across various pre-training strategies in both full-shot and few-shot scenarios show that LEAP consistently outperforms fine-tuning and other prompt-based approaches. \nSpeaker\nMr. Jinfeng Xu\nDepartment of Electrical and Electronic Engineering\nThe University of Hong Kong \nBiography of the Speaker\nJinfeng Xu is a Ph.D. candidate in the Department of Electrical and Electronic Engineering at The University of Hong Kong\, under the supervision of Prof. Edith C. H. Ngai. His current research interests include Recommendation System\, Data Privacy\, Graph Learning\, Self-supervised Learning\, Computer Vision\, and Federated Learning. \nOrganiser\nProf. Edith C. H. Ngai\nDepartment of Electrical and Electronic Engineering\, The University of Hong Kong \nAll are welcome.
URL:https://ece.hku.hk/events/20260130-2/
LOCATION:Online via Zoom
CATEGORIES:Seminar
ATTACH;FMTTYPE=image/jpeg:https://ece.hku.hk/wp-content/uploads/2024/11/rpg-seminar.jpg
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Hong_Kong:20260130T110000
DTEND;TZID=Asia/Hong_Kong:20260130T120000
DTSTAMP:20260511T172843
CREATED:20260123T013325Z
LAST-MODIFIED:20260123T013325Z
UID:114686-1769770800-1769774400@ece.hku.hk
SUMMARY:RPG Seminar – Category Extrapolation for Long-Tail Learning
DESCRIPTION:Zoom Link:\nhttps://hku.zoom.us/j/98761635562?pwd=7QQk1kgMNRmNcaNcVr3bsVi0A2LbZs.1#success \nAbstract\nThis paper tackles long-tailed learning where tail classes suffer from poor feature generalization due to limited sample diversity. It makes a key observation—finer-grained datasets are less harmed by class imbalance—and supports it with quantitative and qualitative evidence showing that increasing granularity improves tail-category feature generalization. Motivated by this\, the authors propose Category Extrapolation: they augment the dataset with open-set\, fine-grained auxiliary classes related to existing categories to strengthen representation learning for both head and tail classes. To automate auxiliary data collection\, they use LLMs as a knowledge base to discover related categories and web crawling to retrieve images. To prevent auxiliary classes from dominating training\, they introduce a neighbor-silencing loss that keeps the model focused on discriminating target classes; at inference time\, auxiliary classifier weights are masked out\, using only the target classes. \nSpeaker\nMr. Shizhen Zhao\nDepartment of Electrical and Electronic Engineering\nThe University of Hong Kong \nBiography of the Speaker\nShizhen Zhao is a PhD student in the Department of Electrical and Electronic Engineering at The University of Hong Kong (HKU). He received the M.S. degree from Huazhong University of Science and Technology and the B.S. degree from Wuhan University of Technology. His research focuses on computer vision\, with particular interests in open-world perception\, long-tail and few-shot learning\, and out-of-distribution (OOD) detection. \nOrganiser\nProf. Xiaojuan Qi\nDepartment of Electrical and Electronic Engineering\, The University of Hong Kong \nAll are welcome.
URL:https://ece.hku.hk/events/20260130/
LOCATION:Online via Zoom
CATEGORIES:Seminar
ATTACH;FMTTYPE=image/jpeg:https://ece.hku.hk/wp-content/uploads/2024/11/rpg-seminar.jpg
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Hong_Kong:20260130T153000
DTEND;TZID=Asia/Hong_Kong:20260130T163000
DTSTAMP:20260511T172843
CREATED:20260121T015143Z
LAST-MODIFIED:20260121T015143Z
UID:114655-1769787000-1769790600@ece.hku.hk
SUMMARY:Seminar on AI Methods for Learning Quantum Systems
DESCRIPTION:Abstract\nAs quantum technologies redefine the landscape of modern electronics and communication networks\, the efficient characterisation and monitoring of complex quantum systems have become essential. In this talk\, I will present AI-driven methods to address the challenges in quantum state learning and property characterisation. First\, I will introduce the Generative Query Network for Quantum (GQNQ)\, which constructs succinct\, data-driven representations from measurement data to accurately predict unseen quantum statistics across diverse quantum systems. I will then discuss a multi-task neural network framework that extracts intricate global properties solely from short-range measurement statistics. Crucially\, our findings demonstrate powerful transfer-learning capabilities\, allowing models trained on small-scale\, classically tractable systems to generalise to much larger quantum systems. These approaches provide a scalable and universal toolkit for optimising next-generation quantum-enhanced electronic and networking systems. \nSpeaker\nDr. Amy Yuexuan WANG \nSpeaker’s Biography\nDr. Amy Yuexuan WANG obtained her Ph.D. from Zhejiang University in 2003. She currently holds a dual appointment as a Senior Research Fellow in the Department of Electrical and Electronic Engineering at The University of Hong Kong (HKU)\, with research expertise in Robotics and Artificial Intelligence\, Wireless Communications and Networking\, Distributed Intelligent Systems and AI for science. \nHer distinguished career includes over 20 years of academic leadership\, with previous roles as a professor at Tsinghua University and as the Associate Director of the AI Lab at HKU. Prof. Wang has an exceptional record in competitive research funding\, having successfully secured and directed numerous major national grants. She has served as the principal investigator for multiple National Key R&D Programs of China and National Natural Science Foundation of China (NSFC) projects\, with a cumulative funding record exceeding RMB 80 million. Her recent applied research focuses on blockchain ecosystem security and privacy-preserving computation\, with technologies deployed in major financial and industrial platforms. \nDr. Wang maintains extensive industry collaborations and holds several patents in areas such as collaborative AI models and 3D printing. Her entrepreneurial work includes co-founding a startup that secured significant contracts for domestically developed AI-CAD software. A recipient of the 2014 National Teaching Achievements Award\, she played a key role in designing Tsinghua University’s pioneering “Yao Class”.  Dr. Wang’s deep technical expertise\, proven leadership in large-scale funded projects\, and strong translational experience will be vital for guiding the project’s technical development and ensuring its practical and commercial impact.
URL:https://ece.hku.hk/events/20260130-1/
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/2026/01/1280-4.jpg
END:VEVENT
END:VCALENDAR