Final Year Projects
I am honored to supervise these innovative final year projects, each a testament to the hard work, creativity, and dedication of my students. Their pioneering ideas continue to inspire and shape the future.
AI For Good
2024
Development of a Mobile Application for Skin Condition Diagnosis and Management
LI, Yuxin; MYAT, Moe Myint
A mobile application is being developed using image recognition technology to diagnose skin conditions, focusing on seven skin diseases (AKIEC, BCC, BKL, DF, MEL, NV, VASC), providing preliminary diagnoses and advice to reduce non-urgent medical consultations and enhance healthcare accessibility.
MindFit: Youth Emotional and Physical Wellbeing Improvement Platform
HO, Chun Tou; HO, Yan Tung
"MindFit" is an application designed to promote mental and physical well-being in youths (aged 15 to 24) through daily check-ins, personalized feedback, and guided exercises to foster healthy habits and coping mechanisms.
2023
Developing Automated Recycling Application with YOLOv8
LI, Wing Lok; NG, Sui Yat
An automated recycling app using YOLOv8 was developed to accurately identify recyclable materials like plastics, glass, aluminum, and paper, with a dirty filter feature enhancing real-world simulation and a user-friendly Android interface.
AI Learning Platform
CHEUNG, Shing Kit; YU, Wai Sum
An educational website has been developed to provide interactive workshops and LLM access, enabling effective learning and utilization of generative AI for solving complex problems through prompt-based learning.
Surgical Counting
2024
Object Recognition for Surgical Counting
FOK, Chung Yan
A deep learning-based computer vision system is being developed to accurately detect and count surgical instruments, optimizing workflow and reducing the risk of retained surgical items (RSI). Through a comparative study of YOLOv10, YOLOv9, and RT-DETRv2, the project focuses on balancing accuracy, speed, and F1-Score, with selected models to be integrated into an embedded system in Term 2.
2023
Enhancing Efficiency in Surgical Instrument Counting: Implementing an Object Recognition System Using YOLOv8
SZE-TO, Kwok Leung; YAU, Hung Kei
An object recognition system using YOLOv8 has been developed to improve surgical instrument counting by providing real-time, accurate identification, outperforming Cascade R-CNN in speed and accuracy. The system, integrated into a full-stack mobile application, demonstrates robustness under low-light and long-distance conditions but requires improvement in handling rotation.