August 2024

Young Scholar TechTalk – Learning Out-of-Distribution Object Detectors from Foundation Models

Innovation Wing Two, TechTalk / By

September 16 2024 (Monday) 4:30-5:30pm
Out-of-distribution (OOD) object detection is a challenging task due to the absence of open-set OOD data. Inspired by recent advancements in text-to-image generative models, such as Stable Diffusion, we study the potential of generative models trained on large-scale open-set data to synthesize OOD samples, thereby enhancing OOD object detection. We introduce SyncOOD, a simple data curation method that capitalizes on the capabilities of large foundation models to automatically extract meaningful OOD data from text-to-image generative models. This offers the model access to open-world knowledge encapsulated within off-the-shelf foundation models. The synthetic OOD samples are then employed to augment the training of a lightweight, plug-and-play OOD detector, thus effectively optimizing the in-distribution (ID)/OOD decision boundaries. Extensive experiments across multiple benchmarks demonstrate that SyncOOD significantly outperforms existing methods, establishing new state-of-the-art performance with minimal synthetic data usage.

SIG – Health AI and Technology (Revita)

LKS Faculty of Medicine, Projects, SIG project / By

Revita is a MedTech initiative under the Health AI and Technology SIG.

We are developing a health monitoring platform designed to enhance healthcare accessibility for seniors by enabling easy use of medical devices. The platform integrates AI to interpret medical data and provides a user-friendly interface that allows anyone to effectively manage and understand senior health needs regardless of their medical training.

TechTalk – Smart Prefabrication Construction: A GBA Case

Innovation Wing Two, TechTalk / By

September 24, 2024 (Tuesday) 4:30-5:30pm
Prefabrication construction is one of the cutting-edge technologies using prefabricated components such as MiC (Modular Integrated Construction) to improve quality, productivity, safety and sustainability. Prefabricated buildings in the Greater Bay Area (GBA), especially those in the Hong Kong Special Administrative Region, have always faced challenges such as high variability, geographical dispersion of activities, and information fragmentation. This talk will take Hong Kong’s prefabricated buildings as an example to demonstrate a smart system implemented using digital twins, big data processing, BIM, and IoT-related technologies. The system uses various captured and collected construction big data for MiC production, logistics, assembly and monitoring. These data will be converted into multi-dimensional information through big data analytics encapsulated in digital clones to map and characterize physical entities such as components, locations, geometries, costs and construction schedules. This research topic is supported by the Innovation and Technology Commission of Hong Kong (ITC) and Chunwo Construction Co., Ltd.

A smart Far UVC system for effective inactivation of surface and airborne bacteria and viruses

Innovation Wing Two, Third exhibition / By

The spread of pathogenic microorganisms in public spaces poses a great threat to human health.

Professor Leung’s team develops a system using far ultraviolet C (UVC) light (wavelength: 222nm) for surface and air disinfection in an actual environment without affecting the normal usage of the area.
Many studies indicated that Far UVC will not create harmful effect on testing creatures such as mice. To further strengthen the safety use of the device for disinfection, the system will not irradiate far UVC light in the presence of people in the area so it will be totally safe in using it.

Young Scholar TechTalk – Next-generation Aqueous Magnesium-ion Batteries

Innovation Wing Two, TechTalk / By

October 8 2024 (Tuesday) 4:30-5:30pm
Aqueous magnesium batteries offer a promising alternative to lithium-ion technology due to their low cost, material abundance, safety, and comparable performance. However, magnesium metal anodes are hindered by passivation, and the narrow electrochemical stability window of aqueous electrolytes significantly limit the battery voltage. My research work introduces innovative aqueous electrolyte systems to address these challenges. A dual-electrolyte magnesium-air battery was developed, achieving a 50% higher peak power density and 46% higher open circuit voltage compared to traditional single-electrolyte systems. Subsequently, a novel water-in-salt electrolyte enabled the first rechargeable aqueous magnesium-ion battery with reversible magnesium metal anode stripping and plating behavior. Furthermore, a quasi-solid-state electrolyte was formulated to regulate ion storage at the cathode, delivering a voltage plateau of 2.6-2.0 V and a remarkable energy density of 264 Wh kg−1, nearly five times higher than current aqueous Mg-ion batteries. This work demonstrates significant advancements in aqueous magnesium batteries, offering a safe and high-performance energy storage solution for a clean energy future.