BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Department of Electrical and Computer Engineering (HKUECE) 電機與計算機工程系 - ECPv6.16.0//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:20230101T000000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTART;TZID=Asia/Hong_Kong:20240405T140000
DTEND;TZID=Asia/Hong_Kong:20240405T150000
DTSTAMP:20260513T102439
CREATED:20240321T081304Z
LAST-MODIFIED:20250114T065546Z
UID:18119-1712325600-1712329200@ece.hku.hk
SUMMARY:RPG Seminar – Diffractive Neural Network Realized by Surface Acoustic Wave System
DESCRIPTION:Speaker\nMr. Lewei HE\nDepartment of Electrical and Electronic Engineering\,\nThe University of Hong Kong \nAbstract\nMatrix-vector multiplication (MVM) is a foundational operation within the architecture of deep neural networks (DNNs)\, critical for the propagation of information between layers and the overall function of the network. Recent advances in computational methodologies have sought to enhance the efficiency of MVM operations\, thereby improving the performance and applicability of DNNs across a spectrum of tasks. One innovative approach to calculate MVM is the utilization of the diffraction process inherent in wave dynamics\, which shares a mathematical resemblance with the operations of MVM.  This conceptual convergence has led to the development of diffractive neural networks\, a novel class of computational systems that employ diffraction phenomena for the execution of MVM tasks. The most common physics system to realize diffractive neural network is optical system suffering the problem of integrated on chip level. Here we propose a novel way of realizing diffractive neural network by surface acoustic wave with high integration level. The seminar will discuss the simulation method of surface acoustic wave diffractive system based on COMSOL. Additionally\, it will illustrate the difference between algorithm of diffractive neural network and tradition neural network. \nBiography of the speaker\nMr. Lewei HE is currently pursuing the MPhil Degree with the Department of Electrical and Electronic Engineering\, The University of Hong Kong (HKU). His research interests focus on simulation of diffractive surface acoustic wave system and algorithm of diffractive neural network. \nOrganizer\nProf. Shiming ZHANG
URL:https://ece.hku.hk/events/20240405-1/
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:20240405T140000
DTEND;TZID=Asia/Hong_Kong:20240405T150000
DTSTAMP:20260513T102439
CREATED:20240327T032308Z
LAST-MODIFIED:20250114T065627Z
UID:18175-1712325600-1712329200@ece.hku.hk
SUMMARY:RPG Seminar – Novel Approaches of Load Redistribution Attacks in Cyber-Physical Power Systems
DESCRIPTION:Abstract:\nWith the increasing integration of renewable energy and related communication technology\, ensuring the cyber-physical security of power systems has become crucial for national industries. In recent times\, hackers have employed load redistribution attacks (LRAs) to evade bad data detection (BDD) systems and manipulate operator actions. In this seminar\, we will discuss innovative LRA approaches in cyber-physical power systems. Firstly\, instead of raising system costs\, we change the LRA’s objective to compromise nodal voltage at specific target locations. Next\, we examine the feasibility of false data injection attacks (FDIAs). Generally\, it is unrealistic for hackers to gain access to network admittance information. To address this\, we create a sequential LRA strategy and estimate the power transfer distribution factor (PTDF) matrix\, allowing for a concealed attack without requiring network admittance knowledge. Lastly\, we expand LRAs to power systems with significant electric vehicle (EV) penetration. We propose a new LRA technique to influence electricity prices in both transmission and distribution networks\, leading to simultaneous charging behavior of EVs and resulting in a load spike during peak hours. \nBiography of the speaker:\nMr. Zelin Liu received his B.Eng. degree in electrical engineering and automation from Zhejiang University and his M.Eng. in electrical engineering from the University of Southern California. Currently\, he is pursuing his Ph.D. in the Department of Electrical and Electronic Engineering at the University of Hong Kong. His primary research interests are power system false data injection attacks and related cyber-physical topics in power systems. \nOrganizer: Prof. Tao LIU
URL:https://ece.hku.hk/events/20240405-2/
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:20240408T143000
DTEND;TZID=Asia/Hong_Kong:20240408T174000
DTSTAMP:20260513T102439
CREATED:20240326T093646Z
LAST-MODIFIED:20250114T065508Z
UID:18170-1712586600-1712598000@ece.hku.hk
SUMMARY:Workshop on Frontiers of Image Science and Visual Computing 2024
DESCRIPTION:For more details\, please visit the event website: https://hku.welight.fun/events
URL:https://ece.hku.hk/events/20240408-1/
LOCATION:Rayson Huang Lecture Theatre\, The University of Hong Kong
CATEGORIES:Seminar
ATTACH;FMTTYPE=image/jpeg:https://ece.hku.hk/wp-content/uploads/2024/03/23423423432.jpg
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Hong_Kong:20240409T110000
DTEND;TZID=Asia/Hong_Kong:20240409T120000
DTSTAMP:20260513T102439
CREATED:20240318T032442Z
LAST-MODIFIED:20250114T065232Z
UID:18082-1712660400-1712664000@ece.hku.hk
SUMMARY:Chiral Meta-photonics with Bound States in the Continuum
DESCRIPTION:Speaker\nProf. Cheng-Wei QIU\nDepartment of Electrical and Computer Engineering\,\nNational University of Singapore \nAbstract\nIn this talk\, the recent progress on how chirality could transcend designs of advanced metasurfaces\, sensing\, BIC photonics and optical micromanipulations will be presented. In particular\, recent results will be reported on the realization of intrinsic chiral metasurfaces where the engineered slant geometry breaks both in-plane and out-of-plane symmetries. The achieved results include intrinsic chiral bound states in the continuum with near-unity CD of 0.93 and quality factor exceeding 2300 for visible frequencies. Finally\, new perspectives on using photonic orbit angular momentum to effectively discriminate single-size and multi-scale chiral nanostructures will be discussed. \nBiography of the Speaker \nProf. Cheng-Wei QIU is appointed IEEE Photonics Society Distinguished Lecturer 2024. He is Fellow of APS\, Optica\, SPIE and The Electromagnetics Academy\, US. He was the recipient of President’s Science Award 2023\, the highest science accolade in Singapore. He was elected Fellow of ASEAN Academy of Engineering and Technology. He is well known for his research in structured light and interfaces. He has published over 480 peer-reviewed journal papers. He was the recipient of URSI Young Scientist Award in 2008\, NUS Young Investigator Award in 2011\, MIT TR35@Singapore Award in 2012\, Young Scientist Award by Singapore National Academy of Science in 2013\, Faculty Young Research Award in NUS 2013\, SPIE Rising Researcher Award 2018\, and Engineering Researcher Award 2021 in NUS\, and World Scientific Medal 2021 by Institute of Physics\, Singapore\, Achievement in Asia Award (Robert T. Poe Prize) by International Organization of Chinese Physicists and Astronomers in 2022. He was Highly Cited Researchers in 2019\, 2020\, 2021\, 2022\, 2023 by Web of Science. He has been serving in Associate Editor for various journals such as JOSA B\, PhotoniX\, Photonics Research\, and Editor-in-Chief for eLight. He also serves in Editorial Advisory Board for Laser and Photonics Review\, Advanced Optical Materials\, and ACS Photonics. \nOrganizer: Prof. Kevin TSIA \nAll are welcome! We look forward to seeing you!
URL:https://ece.hku.hk/events/20240409-1/
LOCATION:Room CB-603\, 6/F\, Chow Yei Ching Building\, The University of Hong Kong
CATEGORIES:Seminar
ATTACH;FMTTYPE=image/jpeg:https://ece.hku.hk/wp-content/uploads/2024/03/20240409-1280.jpg
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Hong_Kong:20240417T100000
DTEND;TZID=Asia/Hong_Kong:20240417T110000
DTSTAMP:20260513T102439
CREATED:20240411T014514Z
LAST-MODIFIED:20250114T065151Z
UID:18254-1713348000-1713351600@ece.hku.hk
SUMMARY:RPG Seminar – Design\, Analysis and Application of Multi-Channel and Multi-Pickup Wireless Power Transfer Systems
DESCRIPTION:Speaker:\nMr. Pang Hongliang\nDepartment of Electrical and Electronic Engineering\,\nThe University of Hong Kong \nAbstract:\nWireless power transfer (WPT) has emerged as a significant technology in modern automation applications. It offers advantages such as reduced installation workload\, enhanced flexibility\, and mobility while eliminating wear and tear associated with traditional power supply cables. In the context of wireless charging\, providing efficient and stable energy supply to multiple load devices presents challenges. These challenges arise due to the coupling effects between loads and variations in the equivalent impedance of batteries. Accordingly\, an impedance buffer-based reactance cancellation method is proposed for a capacitor-inductor-capacitor-series (CLC-S) compensated WPT system with robust zero-voltage-switching (ZVS). The proposed method adopts a synchronous impedance buffer series connected with the CLC compensation network to realize full-range zero-phase-angle (ZPA) operation and can also be regulated to ZVS condition under large tolerance of the abovementioned issues. In addition\, the proposed concept holds the characteristic of communication-free\, feedback-free\, simple control\, and no strict constraints imposed by parameters in high-order compensation.. \nBiography of the speaker:\nMr. Hongliang PANG received the B.Eng. degree in automation and the M.Eng. degree in control science and engineering from Tianjin University\, Tianjin\, in 2017 and 2020\, respectively. He is currently working toward the Ph.D. degree in electrical and electronic engineering with the Department of Electrical and Electronic Engineering\, The University of Hong Kong\, Hong Kong. His current research interests include wireless power transfer and power electronics. \nOrganizer: Prof. Y. HOU \nAll are welcome.
URL:https://ece.hku.hk/events/20240417-1/
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:20240417T163000
DTEND;TZID=Asia/Hong_Kong:20240417T173000
DTSTAMP:20260513T102439
CREATED:20240416T064624Z
LAST-MODIFIED:20250114T065112Z
UID:18270-1713371400-1713375000@ece.hku.hk
SUMMARY:Adaptive Optical Microscopy for Bioimaging
DESCRIPTION:Abstract:\nOptical microscopy has become an indispensable tool for biomedical research\, owing to its noninvasive nature and subcellular resolution. However\, current techniques require further refinement to accurately capture the complexity of living biological systems. In this talk\, I will describe how we integrate ideas from astronomy\, optics\, and computation to address challenges in in vivo imaging. By developing and applying adaptive-optics-assisted microscopy and microendoscopy\, we have achieved high-resolution\, high-contrast\, widefield imaging in living organisms and demonstrated high-resolution two-photon fluorescence imaging of mouse eye pathology and pharmacology in vivo. \nBiography of the Speaker:\nProf. Qinrong ZHANG obtained her Ph.D. degree in Physics from the University of Waterloo. After that\, she joined the University of California\, Berkeley\, as a postdoctoral fellow\, working with Professor Na Ji in the Departments of Physics & of Molecular and Cell Biology. Since August 2023\, she has been an Assistant Professor in the Department of Biomedical Engineering at City University of Hong Kong. Her primary research focus is on developing cutting-edge microscopy technologies for biological studies\, particularly in vivo imaging of the mouse brain and mouse eye. \nOrganizer: Prof. Kevin TSIA \nAll are welcome! We look forward to seeing you!
URL:https://ece.hku.hk/events/20240417-2/
LOCATION:Room CB-208\, 2/F\, Chow Yei Ching\, Building\, The University of Hong Kong
CATEGORIES:Seminar
ATTACH;FMTTYPE=image/jpeg:https://ece.hku.hk/wp-content/uploads/2024/04/324234234.jpg
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Hong_Kong:20240422T140000
DTEND;TZID=Asia/Hong_Kong:20240422T150000
DTSTAMP:20260513T102439
CREATED:20240411T073647Z
LAST-MODIFIED:20250114T064909Z
UID:18267-1713794400-1713798000@ece.hku.hk
SUMMARY:A Bottom-up Approach Towards Generalizable Robot Learning
DESCRIPTION:Abstract:\nThe rise of data-driven methods in robotics has significantly enhanced a robot’s capacity for perception\, reasoning\, and acting. However\, the challenge and expense of collecting a diverse dataset with robots prevent learning control policies that are generalizable across various settings and tasks. Alternatively\, while data sources like videos and robot play data are scalable\, they are often not directly applicable due to the domain gaps and the absence of optimal action labels. In this talk\, I will discuss my research on learning visual representations\, particle trajectory models\, and particle dynamics models from these data to learn generalizable low-level policies. These structured representations enable the learned policies to generalize to novel objects and configurations. I will conclude by demonstrating how these low-level skills can be assembled to tackle long-horizon and novel tasks. \nBiography of the Speaker:\nDr. Xingyu LIN is a postdoctoral researcher at the University of California Berkeley\, working with Pieter Abbeel. His research lies at the intersection of computer vision\, machine learning and robotics\, with a focus on learning robust manipulation skills that generalize to novel objects\, tasks and deformable objects. He holds a PhD from the Robotics Institute at Carnegie Mellon University\, advised by David Held. Prior to that\, he received his undergraduate degree in computer science from Peking University. His research has been published at top conferences\, including CoRL\, RSS\, NeurIPS and ICLR. He was also selected as an RSS (Robotics Science and System) 2022 Pioneer. \nOrganizer: Prof. Kaibin HUANG \nAll are welcome! We look forward to seeing you!
URL:https://ece.hku.hk/events/20240422-2/
LOCATION:Online via Zoom
CATEGORIES:Seminar
ATTACH;FMTTYPE=image/jpeg:https://ece.hku.hk/wp-content/uploads/2024/04/1280-4.jpg
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Hong_Kong:20240422T160000
DTEND;TZID=Asia/Hong_Kong:20240422T180000
DTSTAMP:20260513T102439
CREATED:20240321T060529Z
LAST-MODIFIED:20250114T064821Z
UID:18115-1713801600-1713808800@ece.hku.hk
SUMMARY:Organoid Printing for Precision Cancer Medicine and Regeneration Therapy
DESCRIPTION:Speaker\nDr. Shaohua MA\nAssociate Professor\,\nInstitute of Biopharmaceutical and Health Engineering (iBHE)\,\nTsinghua Shenzhen International Graduate School (SIGS) \nAbstract\nOrganoids are three-dimensional micro-replicas of human organs in of human organs in either physiological or pathological states. Organoids are expected to provide personalised drug treatment choices by assessing drug efficacy and toxicity in efficacy and toxicity in vitro\, or augment regenerative therapy by exploiting their colonized of their colonised stem cell-like capacities. Here we report a bead-jet bead-printing method that is suitable for both reproducible and high-throughput formulation and patterning of organoids. It allows organoid printing in both parallel culture chambers\, such as 96-well plates\, and irregular wounds. It also has significant translation potential by providing an automated tool for organoid and living bead printing. \nBiography of the Speaker\nDr. Shaohua MA received his B.Eng. from the Department of Polymer Materials and Engineering at Sun Yat-sen University in 2009\, and Ph.D. from the Department of Chemistry at the University of Cambridge in 2013. He did postdoctoral training at the University of Oxford in 2013 – 2017 before joining Tsinghua University in 2017\, first as an assistant professor and then as associate professor (core-PI) at Tsinghua-Berkeley Shenzhen Institute (TBSI). He is now an associate professor with tenure at Institute of Biopharmaceutical and Health Engineering (iBHE)\, Tsinghua Shenzhen International Graduate School (SIGS). His research interests include intelligent microfluidics\, organoids and stem cell engineering\, organs-on-a-chip\, 3D bioprinting\, and has contributed over 40 papers as the corresponding author to these fields. \nOrganizer\nDr. Zhiqin CHU \nAll are welcome! We look forward to seeing you!
URL:https://ece.hku.hk/events/20240422-1/
LOCATION:Room CB-601J\, 6/F\, Chow Yei Ching Building\, The University of Hong Kong
CATEGORIES:Seminar
ATTACH;FMTTYPE=image/jpeg:https://ece.hku.hk/wp-content/uploads/2024/03/20240422-1280.jpg
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Hong_Kong:20240423T140000
DTEND;TZID=Asia/Hong_Kong:20240423T150000
DTSTAMP:20260513T102439
CREATED:20240409T091859Z
LAST-MODIFIED:20250114T064707Z
UID:18248-1713880800-1713884400@ece.hku.hk
SUMMARY:Reinforcement Learning with LLMs Interaction for Edge Network-enabled Distributed Diffusion Model
DESCRIPTION:Abstract:\nIn the rapidly evolving field of generative artificial intelligence (GenAI) and AI-generated content (AIGC) services\, generative diffusion models (GDMs) have garnered widespread attention due to their extensive application and exceptional performance\, supporting a series of applications such as Stable Diffusion\, Sora\, and others. Despite their success\, the deployment of GDMs faces significant challenges\, particularly in aligning generated content with individual user preferences and ensuring production efficiency. This presentation introduces a novel user-centric interactive AI approach for edge network-enabled distributed GDM-based AIGC service framework\, prioritizing efficient and collaborative GDM deployment. Specifically\, we restructure the GDM’s inference process\, i.e.\, the denoising chain\, to enable users’ semantically similar prompts to share a portion of diffusion steps. Furthermore\, to maximize the users’ subjective quality-of-experience (QoE)\, we present a reinforcement learning with large language models interaction (RLLI) approach\, which utilizes large language model (LLM)-empowered generative agents to simulate user feedback\, providing real-time and subjective QoE feedback that reflects the spectrum of user personalities. In conclusion\, this presentation seeks to explore the reciprocal relationship between “GenAI for Network” and “Network for GenAI”\, aiming to achieve more efficient\, intelligent\, and sustainable next-generation network services. \nBiography of the Speaker:\nDr. Hongyang DU received a Ph.D. from the School of Computer Science and Engineering\, Energy Research Institute @ NTU\, Nanyang Technological University\, Singapore\, under the Interdisciplinary Graduate Program. He received the B.Sc. degree from Beijing Jiao Tong University\, Beijing\, China\, in 2021. He is the Editor-in-Chief assistant of IEEE Communications Surveys & Tutorials (2022-2024). He was recognized as an exemplary reviewer of the IEEE Transactions on Communications and IEEE Communications Letters in 2021. He was the recipient of the IEEE Daniel E. Noble Fellowship Award from the IEEE Vehicular Technology Society in 2022\, the recipient of the IEEE Signal Processing Society Scholarship from the IEEE Signal Processing Society in 2023\, the recipient of the Chinese Government Award for Outstanding Students Abroad in 2023\, and the recipient of the Singapore Data Science Consortium (SDSC) Dissertation Research Fellowship in 2023. As the team leader\, He won the Honorary Mention award in the ComSoc Student Competition from IEEE Communications Society in 2023 and the First and Second Prizes in the 2024 ComSoc Social Network Technical Committee (SNTC) Student Competition. He has published over 20 first-author papers in leading journals and flagship conferences\, such as the IEEE Journal on Selected Areas in Communications and IEEE Transactions on Mobile Computing. He has also co-authored 100+ peer-reviewed papers with a total of 1\,600+ citations. His research interests include generative artificial intelligence (GenAI)\, edge intelligence\, semantic communications\, and resource allocation. \nOrganizer: Prof. Kaibin HUANG \nAll are welcome! We look forward to seeing you!
URL:https://ece.hku.hk/events/20240423-1/
LOCATION:Online via Zoom
CATEGORIES:Seminar
ATTACH;FMTTYPE=image/jpeg:https://ece.hku.hk/wp-content/uploads/2024/04/1280-3.jpg
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Hong_Kong:20240423T150000
DTEND;TZID=Asia/Hong_Kong:20240423T160000
DTSTAMP:20260513T102439
CREATED:20240409T092103Z
LAST-MODIFIED:20250114T064610Z
UID:18249-1713884400-1713888000@ece.hku.hk
SUMMARY:Have AI Models Truly Simplified Healthcare? Beyond Model-centric AI\, Data- and Human-centric AI for Simplified Healthcare
DESCRIPTION:Abstract:\nIn healthcare\, the exponential growth of artificial Intelligence (AI) has led to the development of numerous advanced AI models for disease diagnosis and lesion segmentation\, etc. Despite these advancements\, the question arises: Have AI models truly simplified healthcare? In this talk\, I will share my thinking on this question\, and present my research aimed at simplifying healthcare in real-world deployment. \nTo effectively simplify healthcare\, we must go beyond just focusing on AI models. Instead\, we should consider the challenges inherent in deploying these models in real clinical practice. These challenges encompass the adaptability of AI models across diverse medical centers with varying medical imaging devices or patient populations (data-centric AI)\, alongside the interaction/collaboration between clinicians and AI models to ensure clinicians’ trust in AI model decisions (human-centric AI). \nAs for data-centric AI\, I have developed many annotation-efficient deep learning technologies to enhance model performance and efficiency when deploying in a new medical scenario\, thus reducing the burden of clinicians in large-scale dataset annotation. On the other hand\, my research on human-centric AI emphasizes the importance of close collaboration and trust between AI and clinicians\, such as enhancing communication between models and clinicians through large language models\, providing more detailed explanations for clinicians’ decision-making processes\, and deferring to radiologists when AI models fail. \nBiography of the Speaker:\nDr. Xiaoqing GUO is a postdoctoral researcher at the Department of Engineering Science\, University of Oxford. She obtained her Ph.D. degree in the Department of Electrical Engineering at the City University of Hong Kong in 2022 and received a B.S. degree from Beihang University in 2018. Her research interest is in the interdisciplinary field of AI and healthcare\, aiming to create innovative intelligent systems that can support high-quality human-machine interaction/collaboration and trustworthy clinical decision-making. In AI and medical imaging fields\, she has published over 30 top journal and conference papers\, including TPAMI\, CVPR\, ICCV\, ECCV\, TMI\, MedIA\, MICCAI\, and Nature\, reaching over 1k Google Scholar citations with an h-index of 15. She has been selected as one of the World’s Top 80 Chinese Young Female Scholars in AI\, and she has received prestigious awards\, such as CVPR Outstanding Reviewer Award\, MSRA Fellowship Nomination Award\, Outstanding Doctoral Research Award\, Outstanding Research Thesis Award\, and three year consecutive Outstanding Academic Performance Award \nOrganizer: Prof. Kaibin HUANG \nAll are welcome! We look forward to seeing you!
URL:https://ece.hku.hk/events/20240423-2/
LOCATION:Online via Zoom
CATEGORIES:Seminar
ATTACH;FMTTYPE=image/jpeg:https://ece.hku.hk/wp-content/uploads/2024/04/1280-2.jpg
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Hong_Kong:20240425T133000
DTEND;TZID=Asia/Hong_Kong:20240425T143000
DTSTAMP:20260513T102439
CREATED:20240409T092329Z
LAST-MODIFIED:20250114T064507Z
UID:18250-1714051800-1714055400@ece.hku.hk
SUMMARY:Small-scale Robots: Miniature but Capable Robots to Revolutionize Healthcare
DESCRIPTION:Abstract:\nWe are witnessing an exponential growth of the impact of small-scale robotics in recent years. These tiny robots have a characteristic length from several millimetres down to hundreds of nanometers. This extreme miniaturization enables these robots to access constrained space buried deep inside human body\, where conventional devices couldn’t reach\, and thus it dramatically benefits minimally invasive diagnosis and treatment. In particular\, millimeter-scale robots are “small enough” to go inside the natural cavities and ducts of human body\, while at the same time they are “big enough” to conduct relatively complex operations\, such as drug delivery and biopsy. In this seminar\, I will introduce my past and current research projects in small-scale robotics\, and discuss my future research plans. I will talk about the design\, fabrication\, control\, and application of these robots. Furthermore\, I will discuss how this emerging field could have imminent and far-reaching positive impact on modern healthcare. \nBiography of the Speaker:\nDr. Jiachen ZHANG is an Assistant Professor at the Department of Biomedical Engineering\, the City University of Hong Kong since 2021. He graduated with a Ph.D. degree from the University of Toronto in 2018. After graduation\, he worked from 2019-2021 as a Humboldt Research Fellow at Max Planck Institute for Intelligent Systems\, Germany. His research focuses on the development of millimeter- and micrometer-scale robotic systems for biomedical applications. He utilizes magnetic field as the primary actuation and control signal for multi-functional small-scale robots. He envisions that small-scale robots are ideal candidates to meet the demands posed by modern healthcare in its evolution to minimize invasiveness. \nOrganizer: Prof. Kaibin HUANG \nAll are welcome! We look forward to seeing you!
URL:https://ece.hku.hk/events/20240425-1/
LOCATION:Room CB-603\, 6/F\, Chow Yei Ching Building\, The University of Hong Kong
CATEGORIES:Seminar
ATTACH;FMTTYPE=image/jpeg:https://ece.hku.hk/wp-content/uploads/2024/04/1280-1.jpg
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Hong_Kong:20240429T100000
DTEND;TZID=Asia/Hong_Kong:20240429T110000
DTSTAMP:20260513T102439
CREATED:20240419T082617Z
LAST-MODIFIED:20250114T064421Z
UID:18341-1714384800-1714388400@ece.hku.hk
SUMMARY:RPG Seminar – Uncertainty Quantification
DESCRIPTION:Abstract:\nUncertainty quantification plays a crucial role in electromagnetic compatibility and inference (EMC/EMI) in the field. Traditionally\, methods such as the Monte Carlo method\, stochastic Galerkin method\, stochastic collocation\, and linear regression have been developed to tackle the challenges of uncertainty quantification (UQ) problems. However\, these methods often face the issue of curse of dimensionality. In this study\, we propose two different approaches to quantify the uncertainty in EMC/EMI for partial equivalent element circuits. The first method utilizes interval analysis to establish the bounds of the quantities of interest. To further capture the stochastic parameters\, we employ the physical-informed neural network to construct the polynomial chaos expansion. Consequently\, the coefficients of the polynomial bases can be obtained. To construct a network without relying on computational models\, we employ the Wasserstein generative adversarial network with a gradient penalty to estimate the stochastic characteristics. This approach allows us to effectively estimate the uncertain properties without explicitly relying on a computational model. \nSpeaker:\nMs. Yuan Ping\nDepartment of Electrical and Electronic Engineering\,\nThe University of Hong Kong \nBiography of the speaker:\nYuan Ping received the B.E. degree and M. S. degree from Xidian University in 2016 and 2019. She is currently pursuing the Ph.D. degree with the Department of Electrical and Electronic Engineering\, The University of Hong Kong\, Hong Kong. Her research interests Uncertainty Quantification\, Phase retrieval and computational electromagnetic. \nOrganizer:   Prof. Lawrence K. YEUNG \nAll are welcome.
URL:https://ece.hku.hk/events/20240429-3/
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:20240429T110000
DTEND;TZID=Asia/Hong_Kong:20240429T120000
DTSTAMP:20260513T102439
CREATED:20240418T011701Z
LAST-MODIFIED:20250114T064341Z
UID:18277-1714388400-1714392000@ece.hku.hk
SUMMARY:RPG Seminar – Enhancing Performances of InGaN-MQW Thin-film Microdisk Laser with Hybrid ODRs
DESCRIPTION:Abstract:\nThe conventional GaN microdisk laser provides poor overlap between the whispering gallery modes (WGM) and the multi-quantum well (MQW) gain region. The thin-film microdisk structure was proposed to overcome this shortcoming\, but the absorptive nature of the metallic bonding layer cum mirror compromises optical confinement. In this work\, a dielectric distributed Bragg reflector (DBR) is integrated with the metallic mirror to form an omni-directional reflector (ODR) that provides high optical reflectance across a wide range of incidence angle to promote optical confinement in the microdisk. Optical-pumped lasing with average lasing threshold power density of 46.5 W/cm2 and Q factors of 18200 (near threshold) is achieved\, representing a major advancement in GaN microdisk laser technology. \nSpeaker:\nMiss Zhongqi WANG\nDepartment of Electrical and Electronic Engineering\nThe University of Hong Kong \nBiography of the speaker:\nMiss Zhongqi WANG received her B.Eng. degree in Materials Science and Engineering at Tsinghua University and now pursuing the Ph.D. degree in the Department of Electrical and Electronic Engineering at the University of Hong Kong. She is now working on GaN based laser fabrication GaN-based microdisk fabrication and analysis of lasing characteristics. \nOrganizer:\nProf. A.H.W. CHOI
URL:https://ece.hku.hk/events/20240429-2/
LOCATION:Room CB-603\, 6/F\, Chow Yei Ching Building\, The University of Hong Kong
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:20240429T140000
DTEND;TZID=Asia/Hong_Kong:20240429T150000
DTSTAMP:20260513T102439
CREATED:20240411T013043Z
LAST-MODIFIED:20250114T064304Z
UID:18251-1714399200-1714402800@ece.hku.hk
SUMMARY:RPG Seminar – Learning A Low-Rank Feature Representation: Achieving Better Trade-Off between Stability and Plasticity in Continual Learning
DESCRIPTION:Speaker:\nMr. Zhenrong LIU\nDepartment of Electrical and Electronic Engineering\,\nThe University of Hong Kong \nAbstract:\nDeep neural networks require the ability to continually learn and adapt to real-world conditions. This ability\, called “continual learning\,” is essential for AI systems to evolve. Among various representative continual learning methods\, null-space-projection-based algorithms have attracted considerable research interest due to their faster training speed and lower memory requirements. These algorithms optimize network parameters in the null space of past tasks’ feature representation matrices\, ensuring stability. However\, quantitatively analyzing the balance between stability and plasticity in null-space-projection-based methods poses significant challenges\, complicating efforts to refine and improve such approaches. \nIn this seminar\, we comprehensively examine null-space-projection-based continual learning methods and uncover two essential insights. Firstly\, to maintain stability\, the rank of the feature covariance increases with the number of continual learning tasks\, leading to a reduction in the dimension of the feature covariance’s null space. Secondly\, the dimension of the feature covariance’s null space significantly influences the plasticity of continual learning. Building on these insights\, we quantitatively demonstrate the stability-plasticity relationship inherent in null-space-projection-based continual learning methods. Then\, based on the stability-plasticity relationship\, we introduce a novel training algorithm named LRFR (Low-Rank Feature Representation) to enhance plasticity without compromising stability. Specifically\, we judiciously select only a subset of neurons in each layer of the network while training individual tasks to learn the past tasks’ feature representation matrix in low-rank. This increases the null space dimension when designing network parameters for subsequent tasks\, thereby enhancing the plasticity. Using CIFAR-100 and TinyImageNet as benchmark datasets for continual learning\, the proposed approach consistently outperforms state-of-the-art methods. \nBiography of the speaker:\nMr. Zhenrong LIU received the B.E. degree in automation from Northeastern University in 2018 and the M.E. degree in information and communication engineering from the Southern University of Science and Technology in 2020. He is currently pursuing the Ph.D. degree with the Department of Electrical and Electronic Engineering\, The University of Hong Kong\, Hong Kong. \nOrganizer: Prof. Yik-Chung WU \nAll are welcome!
URL:https://ece.hku.hk/events/20240429-1/
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:20240430T140000
DTEND;TZID=Asia/Hong_Kong:20240430T150000
DTSTAMP:20260513T102439
CREATED:20240411T072951Z
LAST-MODIFIED:20250114T064121Z
UID:18266-1714485600-1714489200@ece.hku.hk
SUMMARY:RPG Seminar – Communication-Efficient Joint Signal Compression and Activity Detection in Cell-Free Massive MIMO
DESCRIPTION:Speaker:\nMr. Lin Qingfeng\nDepartment of Electrical and Electronic Engineering\,\nThe University of Hong Kong \nAbstract:\nA great amount of endeavour has recently been devoted to device activity detection in massive machine-type communications. This seminar targets at a practical issue: communication-efficient joint signal compression and activity detection in cell-free massive MIMO with capacity-limited fronthauls. To this end\, we propose a novel deep learning framework which jointly optimizes the compression modules\, quantization modules at the access points\, and the decompression module and detection module at the central processing unit. Specifically\, deep unfolding is leveraged for designing the detection module in order to inherit the domain knowledge derived from the optimization algorithm\, and the other modules are constructed by generic layers for increasing the learning capability. A joint training strategy is proposed to optimize all the modules in an end-to-end manner. Numerical results demonstrate the superiority of the proposed end-to-end learning framework compared with classical optimization methods. \nBiography of the speaker:\nMr. Qingfeng LIN received the B.Eng. degree in communication engineering and the M.Eng. degree in information and communication engineering from the Harbin Institute of Technology in 2018 and 2020\, respectively. He is currently pursuing the Ph.D. degree with the Department of Electrical and Electronic Engineering\, The University of Hong Kong\, Hong Kong. \nOrganizer:   Prof. Yik-Chung Wu \nAll are welcome.
URL:https://ece.hku.hk/events/20240430-2/
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:20240430T140000
DTEND;TZID=Asia/Hong_Kong:20240430T150000
DTSTAMP:20260513T102439
CREATED:20240416T065435Z
LAST-MODIFIED:20250114T064217Z
UID:18274-1714485600-1714489200@ece.hku.hk
SUMMARY:RPG Seminar – Towards Parameter-free Ultrasound Localization Microscopy by Vision Transformer
DESCRIPTION:Speaker:\nMr. Wang Renxian\nDepartment of Electrical and Electronic Engineering\,\nThe University of Hong Kong \nAbstract:\nUltrasound localization microscopy (ULM) has emerged as an unprecedented noninvasive microvascular imaging technique that breaks the acoustic diffraction limit. However\, current ULM workflow relies significantly on prior knowledge\, including the impulse response and empirical parameters (e.g.\, the number of microbubbles (MBs) per frame M)\, or training-test dataset consistency in deep learning (DL)-based studies. In this seminar\, a general ULM pipeline that is free from priors will be presented. Specifically\, a channel attention vision transformer model (ViT) was trained using a progressive learning strategy to distill microbubble signals and reduce speckles simultaneously from a single frame without estimation of the impulse response and M. This approach enables the model to learn global information through training on patch sizes that increase progressively. Ample synthetic ultrasound data were generated using the k-Wave toolbox to provide various MB patterns\, thus overcoming the deficiency of labeled data. The ViT output was further processed by a standard radial symmetry method for sub-pixel localization. Our method performed well on model-unseen public datasets: one in silico flow dataset with ground truth and four in vivo datasets of mouse tumor\, rat brain\, rat brain bolus\, and rat kidney\, in terms of precision and accuracy for in silico dataset\, the number of vessels for diverse in vivo datasets while preserving comparable resolutions. The proposed ViT-based model\, seamlessly integrated with state-of-the-art downstream ULM steps\, improved the overall ULM performance with no priors. \nBiography of the speaker:\nMr. Renxian WANG received the B.S. degree in Material Physics from Northwestern Polytechnical University in 2019 and the MPhil degree in Department of Physics from The Chinese University of Hong Kong in 2021\, respectively. He is currently pursuing the Ph.D. degree in the Department of Electrical and Electronic Engineering at the University of Hong Kong\, Hong Kong. \nAll are welcome.
URL:https://ece.hku.hk/events/20240430-1/
LOCATION:Room CB-603\, 6/F\, Chow Yei Ching Building\, The University of Hong Kong
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:20240430T160000
DTEND;TZID=Asia/Hong_Kong:20240430T170000
DTSTAMP:20260513T102439
CREATED:20240422T075807Z
LAST-MODIFIED:20250114T064040Z
UID:18359-1714492800-1714496400@ece.hku.hk
SUMMARY:Towards Human-enabled Intelligent Robots: Perception\, Imitation and Morphology
DESCRIPTION:Meeting ID: 972 6774 1607 \nAbstract:\nThe robotics industry has manufactured multiple successful robots that are deployed in various domains and have been playing a significant role in the modern economy. How to efficiently build\, train and deploy different robots with improved cost and operational safety in diverse tasks in a scalable way? I argue that efficiently using human intelligence embedded in human’s daily activities is the key to help achieve so\, and in this talk\, I will introduce my research works towards achieving this goal. \nI will first introduce my research on extracting useful state information about humans and objects via visual perception\, and focus on efficient training data collection and annotation that can best utilize human capability. I will then introduce my research on human-to-robot imitation\, specifically a new type of methodology that leverages continuous transformation of robot embodiments to co-develop robot hardware and skills\, allowing continuous transformation of a human agent to a robot agent and transferring the human skills along the way. As an application\, I show how this methodology can be used to efficiently control\, design and optimize robots with new morphology and use human experience in this process. I conclude the talk with discussions on my future research plan on improving various aspects of human-enabled safe and low-cost robot systems\, as well as their broader impacts on science\, engineering and society. \nSpeaker:\nDr. Xingyu LIU\nPostdoctoral Associate\,\nCarnegie Mellon University (CMU) \nBiography of the Speaker:\nDr. Xingyu LIU is currently a Postdoctoral Associate at Carnegie Mellon University (CMU) where he works with Professor Ding Zhao in CMU SafeAI Lab. He received his Ph.D. degree from Stanford University where he was advised by Professor Jeannette Bohg. During his Ph.D.\, he spent time in research labs including Google Brain Robotics and Adobe Research. Prior to Ph.D.\, he received M.S. degree from Stanford University and B.Eng. degree from Tsinghua University. His research interest is at the intersection of robotics\, machine learning and computer vision and he reviews regularly for conferences such as RSS\, NeurIPS and CVPR. His research works have been recognized with a Best Paper Award Finalist at CVPR 2022 conference\, a Best Demo Award Finalist at RoboSoft 2024 conference\, multiple (Long) Oral Presentation honors at top AI conferences and are covered by media outlets including Scientific American magazine\, ACM Tech News and O’Reilly. \nOrganizer:\nProf. Kaibin HUANG \nAll are welcome! We look forward to seeing you!
URL:https://ece.hku.hk/events/20240430-3/
LOCATION:Online via Zoom
CATEGORIES:Seminar
ATTACH;FMTTYPE=image/jpeg:https://ece.hku.hk/wp-content/uploads/2024/04/1280.jpg
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Hong_Kong:20240503T140000
DTEND;TZID=Asia/Hong_Kong:20240503T150000
DTSTAMP:20260513T102439
CREATED:20240417T011530Z
LAST-MODIFIED:20250114T063925Z
UID:18275-1714744800-1714748400@ece.hku.hk
SUMMARY:RPG Seminar – Unified Hierarchical Federated Learning: Bridging Autonomous Driving and Construction Inspection
DESCRIPTION:Meeting ID: 957 7820 8166\nPassword: 631839 \nSpeaker:\nMr. Weibin KOU\nDepartment of Electrical and Electronic Engineering\,\nThe University of Hong Kong \nAbstract:\nIn this seminar\, I will present the application of hierarchical federated learning (HFL) to address challenges in two distinct domains: autonomous driving and construction quality defect inspection. For autonomous driving\, we introduce an optimization-based framework\, Communication Resource Constrained Hierarchical Federated Learning (CRCHFL)\, which enhances HFL by incorporating optimization scheme to improve communication efficiency and model generalization under constrained communication resources. The effectiveness of this framework is validated through simulations\, showing significant improvements over traditional federated learning approaches. In the construction sector\, we propose a HFL framework tailored for privacy-preserving collaboration among robots performing quality defect inspections. This method utilizes a lightweight deep learning model suitable for resource-constrained robots\, focusing on image-based crack segmentation to ensure the safety and serviceability of infrastructures. Experimental results demonstrate that this federated approach outperforms the other baselines. Both implementations underline the versatility and efficiency of HFL in processing large datasets across distributed environments while adhering to privacy constraints\, offering substantial improvements in both operational efficiency and data security. \nBiography of the speaker:\nMr. Weibin KOU is currently working toward a Ph.D. degree with the Department of Electrical and Electronic Engineering\, The University of Hong Kong. His research interests include federated learning\, autonomous driving and robotic perception\, and Large Models (LMs). \nOrganizer:\nProf. Yik-Chung WU \nAll are welcome.
URL:https://ece.hku.hk/events/20240503-1/
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:20240506T150000
DTEND;TZID=Asia/Hong_Kong:20240506T160000
DTSTAMP:20260513T102439
CREATED:20240429T093019Z
LAST-MODIFIED:20250114T063840Z
UID:18470-1715007600-1715011200@ece.hku.hk
SUMMARY:RPG Seminar – On-the-fly communication-and-computing for distributed data analytics and edge intelligence
DESCRIPTION:Abstract:\nEnormous amounts of data are generated by billions of edge devices in mobile networks. Distributed data analytics can support a broad range of mobile applications\, from edge AI to IoT sensing. Enabling such analytics while improving its effectiveness has triggered a paradigm shift from separated optimization between communication techniques and computation algorithms to a joint design. \nConventionally\, the wireless implementation of computation algorithms\, such as statistic data analytics and AI models\, has followed a one-shot approach. This approach first computes local results at devices using local data and then aggregates them to a server with communication-efficient techniques. However\, this implementation is confronted with issues such as limited on-device storage and computation capacities\, link interruption\, and coarse efficiency-accuracy trade-offs. \nIn this seminar\, I will introduce a novel framework of on-the-fly communication-and-computing (FlyCom2). FlyCom2 exploits streaming low-complexity computation and progressive transmission to realize demanding computation algorithms in a mobile network\, such as distributed data analytics and device-server fine-tuning of large language models (LLMs). I will elaborate on the distinct features and advantages of FlyCom2 as well as the possible challenges for materializing it. Furthermore\, I will introduce two use cases explored in my studies on FlyCom2. \nSpeaker:\nMr. Xu CHEN\nDepartment of Electrical and Electronic Engineering\,\nThe University of Hong Kong \nBiography of the speaker:\nMr. Xu CHEN received the B.Eng. and M.Eng. from Harbin Institute of Technology (HIT)\, Harbin\, China in 2018 and 2020\, respectively. He is currently pursuing the Ph.D. degree with the Department of Electrical and Electronic Engineering\, The University of Hong Kong\, Hong Kong. His research interests include MIMO communications\, distributed computing\, and integrated sensing and edge AI. \nOrganizer:\nProf. Kaibin HUANG \nAll are welcome.
URL:https://ece.hku.hk/events/20240506-2/
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:20240506T160000
DTEND;TZID=Asia/Hong_Kong:20240506T170000
DTSTAMP:20260513T102439
CREATED:20240429T092539Z
LAST-MODIFIED:20250114T063806Z
UID:18469-1715011200-1715014800@ece.hku.hk
SUMMARY:RPG Seminar – Vertical Layering of Quantized Neural Networks for Heterogeneous Inference
DESCRIPTION:Abstract:\nAlthough considerable progress has been obtained in neural network quantization for efficient inference\, existing methods are not scalable to heterogeneous devices as one dedicated model needs to be trained\, transmitted\, and stored for one specific hardware setting\, incurring considerable costs in model training and maintenance. In this seminar\, we study a new vertical-layered representation of neural network weights for encapsulating all quantized models into a single one. It represents weights as a group of bits (i.e.\, vertical layers) organized from the most significant bit (also called the basic layer) to less significant bits (i.e.\, enhance layers). Hence\, a neural network with an arbitrary quantization precision can be obtained by adding corresponding enhance layers to the basic layer. However\, we empirically find that models obtained with existing quantization methods suffer severe performance degradation if they are adapted to vertical-layered weight representation. To this end\, we propose a simple once quantization-aware training (QAT) scheme for obtaining high-performance vertical-layered models. Our design incorporates a cascade downsampling mechanism with the multi-objective optimization employed to train the shared source model weights such that they can be updated simultaneously\, considering the performance of all networks. After the model is trained\, to construct a vertical-layered network\, the lowest bit-width quantized weights become the basic layer\, and every bit dropped along the downsampling process act as an enhance layer. Experiments show that the proposed vertical-layered representation and developed once QAT scheme are effective in embodying multiple quantized networks into a single one and allow one-time training\, and it delivers comparable performance as that of quantized models tailored to any specific bit-width. \nSpeaker:\nMr. Hai WU\nDepartment of Electrical and Electronic Engineering\,\nThe University of Hong Kong \nBiography of the speaker:\nMr. Hai WU (Graduate Student Member\, IEEE) received the BEng degree from the Department of Electronic and Electrical Engineering\, Southern University of Science and Technology\, China\, in 2020. He is currently working toward the PhD degree with the Department of Electrical and Electronic Engineering\, The University of Hong Kong\, Hong Kong. \nOrganizer:\nProf. Kaibin HUANG \nAll are welcome.
URL:https://ece.hku.hk/events/20240506-1/
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:20240507T100000
DTEND;TZID=Asia/Hong_Kong:20240507T110000
DTSTAMP:20260513T102439
CREATED:20240426T084308Z
LAST-MODIFIED:20250114T063723Z
UID:18465-1715076000-1715079600@ece.hku.hk
SUMMARY:RPG Seminar – Hybrid Module with Multiple Receptive Fields and Self-attention Layers for Medical Image Segmentation
DESCRIPTION:Meeting ID: 958 6149 4641\nPassword: 505358 \nAbstract:\nRecent advances in medical image segmentation models combine convolution with the attention mechanism which provides an effective approach to formulate long-term dependencies. However\, many works either replaced the convolutional layers with attention layers or embedded attention layers into convolutional neural network (CNN)-based models. To explore the potential of hybrid architecture\, we propose a simple cascade module that builds up multiple receptive fields using convolutional kernels with different sizes and learns global context via self-attention layers. Benefiting from the powerful representation ability of the proposed module\, multilayer perceptrons (MLPs) with shift operation are adopted to bridge the encoder and decoder to reduce the model size without losing accuracy. Experiments show that our model consistently outperforms the latest 2D and 3D models by large margins on three public tasks and is more resilient to shape\, size\, and boundary variations. \nSpeaker:\nMr. Wenbo QI\nDepartment of Electrical and Electronic Engineering\nThe University of Hong Kong \nBiography of the speaker:\nMr. Wenbo QI received the B.Eng. degree from the University of Science and Technology of China in 2019\, and the M.Eng. degree from The University of Hong Kong in 2020\, where he is currently pursuing the Ph.D. degree with the Department of Electrical and Electronic Engineering. His research interests include computer vision\, medical image processing. \nOrganizer:\nProf. S. C. CHAN \nAll are welcome.
URL:https://ece.hku.hk/events/20240507-1/
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:20240508T100000
DTEND;TZID=Asia/Hong_Kong:20240508T110000
DTSTAMP:20260513T102439
CREATED:20240429T062039Z
LAST-MODIFIED:20250114T063647Z
UID:18467-1715162400-1715166000@ece.hku.hk
SUMMARY:RPG Seminar – A New Adaptive Fading Instrumental Variable Pseudolinear Kalman Filter for 3D AOA Target Tracking
DESCRIPTION:Meeting ID: 990 0206 5927\nPassword: 585304 \nAbstract:\nThe instrumental variable pseudolinear Kalman filter (IV-PLKF) algorithm\, used for 3D angle-of-arrival (AOA) target tracking\, has been proven to be more robust to initialization errors\, with superior estimation performance and lower computational complexity compared to other state-of-the-art methods. However\, the IV-PLKF algorithm requires prior knowledge of the state and angle measurement noise information\, which is not available in practice. Improper selection of these values or mismatches due to time-varying changes can significantly impact the stability and estimation performance of the algorithm. To address this issue\, we propose a new adaptive fading (AF-) IV-PLKF algorithm that adaptively mitigates the possible scale mismatches in the state and measurement noise covariance matrices and the IV parameters. Simulation results demonstrate that the proposed algorithm outperforms the conventional IVPLKF under mismatched state and measurement noise covariance scenarios. Moreover\, the proposed method can even achieve comparable estimation performance to that of IV-PLKF with perfect knowledge of the noise information. \nSpeaker:\nMs. Mengxia HE\nDepartment of Electrical and Electronic Engineering\,\nThe University of Hong Kong \nBiography of the speaker:\nMs. Mengxia HE received her B.Eng. degree from the University of Science and Technology Beijing in 2018 and her M.Eng. degree from the Beijing University of Posts and Telecommunications in 2021. She is currently pursuing the Ph.D. degree with the Department of Electrical and Electronic Engineering\, The University of Hong Kong\, Hong Kong. \nOrganizer:\nProf. S. C. CHAN \nAll are welcome.
URL:https://ece.hku.hk/events/20240508-1/
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:20240508T103000
DTEND;TZID=Asia/Hong_Kong:20240508T113000
DTSTAMP:20260513T102439
CREATED:20240429T062517Z
LAST-MODIFIED:20250114T063611Z
UID:18468-1715164200-1715167800@ece.hku.hk
SUMMARY:RPG Seminar – Transformer-based Architectures for Automated Annotation in 3D Point Clouds
DESCRIPTION:Abstract\nManual annotation of 3D point clouds is notoriously labor-intensive\, prompting the need for automated solutions. Existing automated annotation methods\, however\, are typically complex and may neglect the crucial inter-object feature relationships that are informative for annotating challenging samples. In response\, we introduce two end-to-end Transformer-based models\, CAT and CAT++\, which are streamlined to serve as automated 3D-box labelers. These models leverage a minimal set of human annotations to produce precise 3D box annotations from 2D boxes. Our architecture employs a dual encoder strategy: a local intra-object encoder and a global inter-object encoder\, both utilizing self-attention mechanisms to process sequence and batch dimensions. The intra-object encoder captures point-level interactions within objects\, while the inter-object encoder discerns feature relationships across objects\, enhancing scene comprehension. The advanced CAT++ model incorporates a Hierarchical-interleaved encoding scheme and an implicit neural representation\, further refining the annotation process. Benchmarking experiments on the KITTI and nuScenes datasets demonstrate our models’ superior performance over current state-of-the-art methods\, particularly in annotating complex scenarios encompassing all hard samples. \nSpeaker\nMs. Xiaoyan QIAN\nDepartment of Electrical and Electronic Engineering\,\nThe University of Hong Kong \nBiography of the speaker\nMs. Xiaoyan QIAN received the B.Eng. degree in Industrial Engineering from the Zhejiang University of Technology\, Zhejiang\, China. She is currently a Ph.D. candidate in the Department of Electrical and Electronic Engineering at the University of Hong Kong\, under the supervision of Dr. N Wong and Prof. SC Tan. Her current research interests mainly focus on 3D point clouds\, weakly supervised 3D object detection\, and auto-driving. \nOrganizer\nProf. N. WONG \nAll are welcome.
URL:https://ece.hku.hk/events/20240508-2/
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:20240514T140000
DTEND;TZID=Asia/Hong_Kong:20240514T150000
DTSTAMP:20260513T102439
CREATED:20240503T092811Z
LAST-MODIFIED:20250114T063407Z
UID:18489-1715695200-1715698800@ece.hku.hk
SUMMARY:RPG Seminar – Complex-valued Transformer for Wireless Communications
DESCRIPTION:Meeting ID: 892 6022 4678\nPassword: n9QxYg \nAbstract\nIn recent years\, attention-based models\, particularly those employing the transformer structure\, have exhibited exceptional performance in tasks such as natural language processing\, computer vision\, and wireless communication\, among others. Notably\, most of these transformer frameworks rely on real-valued operations\, where inputs\, outputs\, and trainable parameters are real numbers\, even in tasks involving complex-valued domains.\nComplex-valued neural networks (CVNN) have emerged as a powerful approach for addressing problems associated with the complex-valued nature of input data. However\, the application of complex-valued transformers remains largely unexplored within the field of wireless communication\, where most task inputs\, such as received symbols and channel coefficients\, are intrinsically complex-valued. \nThis seminar aims to unveil the potential of complex-valued transformers for complex-valued domain tasks in wireless communications\, specifically focusing on channel estimation in single-input-single-output system and device activity detection in grant-free massive access scenario. To this end\, we propose tailored complex-valued transformer designs that incorporate complex-valued attention mechanisms for both tasks. In particular\, the proposed frameworks exploit the relationship between the real and imaginary parts of signals as implicit constraints\, while capturing temporal and spatial correlations of complex-valued input features. Moreover\, we introduced a novel complex-to-real layer to convert the complex-valued feature into the probabilistic representation for the activity detection task. \nNumerical results demonstrate the superiority of the proposed complex-valued transformers framework compared with other deep-learning based methods and optimization approaches. \nSpeaker\nMiss Leng Yang\nDepartment of Electrical and Electronic Engineering\,\nThe University of Hong Kong \nBiography of the speaker\nMiss Leng Yang received the B.Eng. degree in Electronic and Information Science and Technology from Fudan University in 2022. She is currently pursuing the MPhil degree with the Department of Electrical and Electronic Engineering\, The University of Hong Kong\, Hong Kong. \nOrganizer\nProf. Yik-Chung WU \nAll are welcome.
URL:https://ece.hku.hk/events/20240514-1/
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:20240514T140000
DTEND;TZID=Asia/Hong_Kong:20240514T150000
DTSTAMP:20260513T102439
CREATED:20240503T094824Z
LAST-MODIFIED:20250114T063447Z
UID:18490-1715695200-1715698800@ece.hku.hk
SUMMARY:RPG Seminar – Image Augmented Multimodal Autolabeller for 3D Object Detection
DESCRIPTION:Abstract\nRecently deep learning methods have gained groundbreaking success in many areas\, including autonomous driving and 3D object detection. Powerful neural networks are proposed and yield human-comparable ability after being trained from large datasets. Nonetheless\, the annotation procedure is time-consuming and tedious. To automate the annotation process\, we proposed two methods called MAP-Gen and MTrans\, respectively. Leveraging both image and point cloud modalities\, the two methods can effectively alleviate the sparsity problem of point clouds and hence generate high-quality pseudo labels. \nSpeaker\nMr. Liu Chang\nDepartment of Electrical and Electronic Engineering\,\nThe University of Hong Kong \nBiography of the speaker\nMr. Liu Chang received his B.Eng. degree in Computer Engineering from the University of Hong Kong. He is currently a Ph.D. student supervised by Dr. N. Wong and Prof. Edmund Y. Lam\, at the Department of Electrical and Electronic Engineering\, University of Hong Kong. His current research interests include 3D Vision\, Point Cloud\, Multi-modal Neural Networks\, and NLP. \nOrganizer\nProf. N. WONG
URL:https://ece.hku.hk/events/20240514-2/
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:20240514T140000
DTEND;TZID=Asia/Hong_Kong:20240514T150000
DTSTAMP:20260513T102439
CREATED:20240506T013049Z
LAST-MODIFIED:20250114T063519Z
UID:18492-1715695200-1715698800@ece.hku.hk
SUMMARY:RPG Seminar – Memristor Enabling Efficient Combinatorial Optimization with Quantum-inspired Parallel Annealing
DESCRIPTION:Abstract\nCombinatorial optimization problems are prevalent in various fields\, but obtaining exact solutions remains challenging due to the combinatorial explosion with increasing problem size. Special-purpose hardware such as Ising machines\, particularly memristor-based analog Ising machines\, have emerged as promising solutions. However\, existing simulate-annealing-based \nimplementations have not fully exploited the inherent parallelism and analog storage/processing features of memristor crossbar arrays. This work proposes a quantum-inspired parallel annealing method that enables full parallelism and improves solution quality\, resulting in significant speed and energy improvement when implemented in analog memristor crossbars. We experimentally solved tasks\, including unweighted and weighted Max-Cut and traveling salesman problem\, using our integrated memristor chip. The quantum inspired parallel annealing method implemented in memristor-based hardware has demonstrated significant improvements in time- and energy- efficiency compared to previously reported simulated annealing and Ising machine implemented on other technologies. This is because our approach effectively exploits the natural parallelism\, analog conductance states\, and all-to-all connection provided by memristor technology\, promising its potential for solving complex optimization problems with greater efficiency. \nSpeaker\nMr. Mingrui JIANG\nDepartment of Electrical and Electronic Engineering\,\nThe University of Hong Kong \nBiography of the speaker\nMr. Mingrui JIANG received the B.E. degree from School of Optical and Electronic Information\, Huazhong University of Science and Technology\, Wuhan\, China\, in 2020. He is currently pursuing the Ph.D. degree with the Department of Electrical and Electronic Engineering\, The University of Hong Kong\, Hong Kong SAR. His research interests include analog signal processing\, analog in-memory computing and neuromorphic computing based on emerging memory devices (e.g.\, memristors). \nOrganizer\nProf. Can LI \nAll are welcome.
URL:https://ece.hku.hk/events/20240514-3/
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:20240515T160000
DTEND;TZID=Asia/Hong_Kong:20240515T170000
DTSTAMP:20260513T102439
CREATED:20240304T071147Z
LAST-MODIFIED:20250114T072111Z
UID:17971-1715788800-1715792400@ece.hku.hk
SUMMARY:Microneedle-based Skin Patch for Transdermal Drug Delivery and Biosensing
DESCRIPTION:Microneedles (MNs) are an emerging platform for transdermal applications including drug delivery\, insulin delivery\, vaccination\, biosensing\, disease diagnosis\, and cosmeceutical industry. Their advantages lie in their easy-to-use\, pain-free\, minimally invasive\, and self-administrable features. This overcomes the skin barrier to enhance transdermal delivery of drugs and biomolecules with different physicochemical properties in vitro\, ex vivo and in vivo. In this talk\, Prof. Xu will share microneedle technologies developed in his lab for meeting a wide range of medical needs\, including keloid treatment and prevention\, obesity treatment\, dental and eye disease treatment\, and immune therapies. He will also present his envision in utilizing MN platform for the in-situ monitoring of physiological signals. \nBiography of the speaker: \nProf. Chenjie XU got his PhD\, Master\, and BS from Brown University (2009)\, HKUST (2004)\, Nanjing University (2002) respectively. He had conducted research at Stanford University (2005)\, Brigham and Women’s Hospital (2009-2012)\, and Nanyang Technological University (2012-2019). Currently\, he is an associate professor of biomedical engineering at the City University of Hong Kong. Prof. XU is dedicated to the development of transdermal drug delivery formulations and devices (especially nucleic acid-based nanoparticles and microneedle-based skin patches). He is well known for the development of skin patch for keloid treatment\, anti-obese skin patch\, skin patch for skin interstitial fluid extraction etc. He has published more than 140 peer-reviewed articles (citation is 11k with H index of 45)\, edited two books\, holding 10 international patents\, and found two spin-offs. His research is supported by a wide range of public and private foundations including Singapore Minister of Education\, Singapore A*Star\, Continental Corp (German)\, Bill & Melinda Gates Foundation\, Hong Kong University Grants Committee\, National Natural Science Foundation of China\, etc. \nAll are welcome
URL:https://ece.hku.hk/events/20240315/
LOCATION:Room CB-601J\, 6/F\, Chow Yei Ching Building\, The University of Hong Kong
CATEGORIES:Seminar
ATTACH;FMTTYPE=image/jpeg:https://ece.hku.hk/wp-content/uploads/2024/03/20240315-banner.jpg
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Asia/Hong_Kong:20240516T110000
DTEND;TZID=Asia/Hong_Kong:20240516T120000
DTSTAMP:20260513T102439
CREATED:20240507T081518Z
LAST-MODIFIED:20250114T063319Z
UID:18499-1715857200-1715860800@ece.hku.hk
SUMMARY:RPG Seminar – Domain-Specific Efficient Neural Network Architecture Design
DESCRIPTION:Abstract\nAI models significantly impact our daily lives\, but their high performance brings the challenge of model complexity. Deploying these models on edge devices poses additional challenges\, including power consumption\, memory storage and latency constraints. In this seminar\, we will delve into designing efficient neural network architectures for various domains\, including low-level computer vision and neural fields. We will start by discussing the latest Lookup Table (LUT)-based approach for Single-Image Super-Resolution (SISR). Our proposed Hundred-Kilobyte LUT (HKLUT) requires only 100KB\, 10X less than the second smallest LUT-based method\, and delivers superior performance. Moreover\, we will explore the field of Implicit Neural Representation (INR)\, where inference efficiency is often overlooked. We propose the Activation-Sharing Multi-Resolution (ASMR) coordinate network to enhance INR’s rendering efficiency. By sharing activations across data grids\, ASMR can reduce its Multiply-Accumulate (MAC) operations by up to 500x and improve reconstruction quality. \nSpeaker\nMr. Jason Chun Lok LI\nDepartment of Electrical and Electronic Engineering\,\nThe University of Hong Kong \nBiography of the speaker\nMr. Jason Chun Lok LI holds a BEng degree from the Department of Electrical and Electronic Engineering at The University of Hong Kong\, obtained in 2020. He is currently continuing his studies at the same institution\, working towards a PhD. His research interest lies in the development of domain-specific techniques for efficient deep learning on edge devices. \nOrganizer\nProf. Ngai WONG \nAll are welcome.
URL:https://ece.hku.hk/events/20240516-1/
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:20240516T143000
DTEND;TZID=Asia/Hong_Kong:20240516T153000
DTSTAMP:20260513T102439
CREATED:20240507T082448Z
LAST-MODIFIED:20250114T063241Z
UID:18500-1715869800-1715873400@ece.hku.hk
SUMMARY:RPG Seminar – Exploration of Novel Operators with Memristor Arrays Towards Efficient and Robust In-memory Computing
DESCRIPTION:Abstract\nThe past decade of escalated development in deep learning (DL) has achieved unprecedented success in engineering fields. In particular\, deep neural networks (DNNs) via deep learning have achieved remarkable success across various applications. However\, challenges remain in the hardware implementation of these software-oriented AI algorithms\, primarily due to the reliance on traditional von Neumann computing architectures which are inefficient and lead to high power usage and latency particularly at the edge computing level. To address these issues\, compute-in-memory (CIM) using non-volatile memristive devices presents a promising solution. CIM leverages in-memory data processing to reduce data movement\, thereby improving efficiency. Therefore\, a core issue in artificial intelligence-related fields lies in leveraging hardware practice experience to explore and develop neuron models and operational operators. In the upcoming talk\, an innovative memristive unit cell based on the arithmetic unit model will be introduced\, aiming to explore its performance and robustness in emerging operational networks within AI fields. \nSpeaker\nMr. Yuan REN\nDepartment of Electrical and Electronic Engineering\,\nThe University of Hong Kong \nBiography of the speaker\nMr. Yuan REN received the M.S. degree in electrical and computer engineering from the University of Macau (UM)\, Macao. He then joined the SoC Key Laboratory\, Peking University Shenzhen Institute and PKU-HKUST Shenzhen-Hong Kong Institution\, Guangdong\, China. He is currently pursuing the Ph.D. degree in electrical and electronic engineering from the University of Hong Kong (HKU)\, under the supervision of Dr. Ngai Wong. His research focuses on algorithm-hardware co-design for AI accelerator and memristor-based compute-in-memory integrated circuits. \nOrganizer\nProf. Ngai WONG \nAll are welcome.
URL:https://ece.hku.hk/events/20240516-2/
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:20240517T100000
DTEND;TZID=Asia/Hong_Kong:20240517T110000
DTSTAMP:20260513T102439
CREATED:20240510T012737Z
LAST-MODIFIED:20250114T063205Z
UID:18502-1715940000-1715943600@ece.hku.hk
SUMMARY:RPG Seminar – Manipulating Light Scattering at the Nanoscale by Metasurface
DESCRIPTION:Abstract\nLight scattering is a fundamental optical process that accounts for many optical phenomena and applications. This process comes from the interaction between light and scattering particles\, or scatters. It greatly depends on parameters such as the scatters’ shapes and refractive index\, the polarization and wavelength of light. We will show that by arranging the specially designed nano scatters on a flat surface to form a metasurface\, the output light field can be manipulated at the nanoscale\, which will lead to many promising applications. \nTwo main topics will be discussed in this seminar. The first topic relates to tri-channel metalenses. Since it is difficult to encode three independent phase information at single-pixel or single-cell level\, most current designs use spatial multiplexing strategies including segmentation\, interleaving and multilayer integration\, which would result in large unit pixel sizes and limited performances. In this seminar\, we will present a single-celled design method to achieve tri-functional metalenses. Another topic relates to broadband antireflection by metasurfaces. We have proposed a quasi-random design method\, and developed a high-throughput nanofabrication method to fabricate the metasurfaces. \nSpeaker\nMr. Xudong GUO\nDepartment of Electrical and Electronic Engineering\,\nThe University of Hong Kong \nBiography of the speaker\nMr. Xudong GUO received the B.Eng. degree in Optoelectronic Information Science and Engineering from Changchun University of Science and Technology\, Changchun\, in 2018. He is currently working toward the Ph.D. degree in electrical and electronic engineering with the Department of Electrical and Electronic Engineering\, The University of Hong Kong\, Hong Kong. His research interests include metasurface\, holography and imaging. \nOrganizer\nProf. Kenneth K. Y. WONG
URL:https://ece.hku.hk/events/20240517-2/
LOCATION:Room CB-603\, 6/F\, Chow Yei Ching Building\, The University of Hong Kong
CATEGORIES:Seminar
ATTACH;FMTTYPE=image/jpeg:https://ece.hku.hk/wp-content/uploads/2024/11/rpg-seminar.jpg
END:VEVENT
END:VCALENDAR