Seminar

Meeting ID: 957 7820 8166 Password: 631839 Speaker: Mr. Weibin KOU Department of Electrical and Electronic Engineering, The University of Hong Kong Abstract: In 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 […]

Abstract: Enormous 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 […]

Abstract: Although 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 […]

Meeting ID: 958 6149 4641 Password: 505358 Abstract: Recent 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 […]

Meeting ID: 990 0206 5927 Password: 585304 Abstract: The 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 […]

Abstract Manual 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 […]

Meeting ID: 892 6022 4678 Password: n9QxYg Abstract In 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 […]

Abstract Recently 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, […]

Abstract Combinatorial 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 implementations have not fully exploited the inherent parallelism and analog storage/processing features […]

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 […]

Abstract AI 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 […]

Abstract The 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 […]

Abstract Light 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 […]

Abstract In the past decades, with the rapidly increasing data, AI technology, including neural network (NN) shows more and more powerful ability. However, the development of electronic hardware meets a dilemma because of the physical limitation, which restricts the computation performance growth of NN. Thus, developing the next-generation computation platform for NN is necessary. Since […]

Abstract: Blood analysis is an indispensable clinical tool for human health and diseases. Specifically, the overarching challenge in characterization of blood, notably immune cells, is to identify the cellular phenotypes at the single-cell precision to dissect the complex functional roles of different cell types/states. Strategies for phenotyping immune cells enable biological discovery and shed light […]

Abstract: Financial portfolio management (PM) is a very important topic in computational finance, with its primary objective of achieving higher returns while reducing investment risks through dynamically allocating capital to different assets in a portfolio. Recently, deep or reinforcement learning(DL/RL)-based PM approaches have been applied to capture the valuable opportunities from the underlying financial market […]

Abstract Layered two-dimensional (2D) materials interact primarily via van der Waals bonding, which has created new opportunities for heterostructures that are not constrained by epitaxial lattice matching requirements. However, since any passivated, dangling bond-free surface interacts with another via non-covalent forces, van der Waals heterostructures are not limited to 2D materials alone. In particular, 2D […]

Abstract Despite the long and crucial role of traditional solid-state physics for current silicon-based technologies, next-generation neuromorphic, non-volatile memory, and energy devices that are key components in the era of the Internet of Things (IoT) require novel working principles with quantum physics emerging in low-dimensional materials. The main research direction for future devices is to […]

Abstract The need to shuttle data between processing and memory units has been a key performance bottleneck for conventional hardware implementations of artificial neural networks. Analog in-memory computing (AIMC) emerges as a promising solution to this challenge by directly performing computations within non-volatile memory devices, such as phase change memory (PCM). However, the utilization of […]

The Faculty of Engineering, The University of Hong Kong is organizing the "Workshop on Frontiers of Legged Robotics 2024". The workshop will be held on May 23, 2024 (Thursday) from 2:30 pm to 6:00 pm at Lecture Theatre A, G/F, Chow Yei Ching Building, HKU. Keynote Speaker for William Mong Distinguished Lecture: - Prof Abderrahmane […]

Abstract As the penetration of variable renewable generation increases in power systems, issues such as grid stiffness, larger frequency deviations, and grid stability are becoming more relevant. In this context, Energy Storage Systems (ESSs) are proving to be effective in facilitating the integration of renewable resources, and thus are being widely deployed in both microgrids […]

Abstract Position-related services and applications appear everywhere in our daily lives and precise positioning is indispensable.  For non-cooperative localization of an object by time difference of arrival (TDOA) such as in an integrated sensing and communications (ISAC) environment, it requires the knowledge that the object is near to or far from the receivers.  If the […]

Meeting ID: 939 1846 8563 Abstract Optical nanostructures and two-dimensional materials (2DM) have optical properties that are widely tunable via several approaches, such as heating, electrostatic gating, and interfacial engineering such as twisting. Being able to tailor the interfacial properties in a similar real-time manner represents the next leap in our ability to modulate the underlying […]

Meeting ID:937 8072 8049 Abstract Photonics technology has become indispensable in communication, imaging, sensing, metrology, and more recently, in signal processing and computation. These advancements are made possible through optical systems that leverage the control of multiple degrees of freedom of light, e.g., in space/momentum, time/frequency, angle/orbital angular momentum and polarization. Recently, the integration of […]