Capstone Projects Highlight

The Department provides students with the opportunities to apply and integrate their knowledge acquired throughout the undergraduate study. In their final year, students are required to develop a Capstone Project by analyzing and solving complex and possibly real-life problems with skills on systematic development and documentation of a significant piece of work.

eHealth Cloud System

Student: CHUI Cheuk Long

Objective: 

- To relieve the rise of aging population
- To improve the current implementation of cloud healthcare system which requires rigid and large cloud infrastructure
- To propose a newly designed low cost IoT -> AI + Micro Cloud pipeline

Sentiment Classification of Cantonese Restaurant Reviews UPOW: The Instant Finality Consensus Protocol

Student: JIAO Ying

Objective: 

- To obtain a Cantonese corpus
- To construct a Cantonese sentiment lexicon
- To investigate approaches of integrating sentiment lexicon into deep learning to conduct Cantonese sentiment classification

UPOW: The Instant Finality Consensus Protocol

Student: LEE Chi Yin

Objective: To introduce new consensus protocol to achieve instant finality, high throughput and responsiveness blockchain in permission-less network for decentralised e-cash system..

Blockchain-Based Closed-Loop Plastic Bottle Recycle System  

Student: LEE Chuck Lun Edward

Objective: To solve the real-world problems related to plastic bottles including over spilling bin, uncategorised trash, bottles to landfill by developing a closed-loop system that benefits to the public.

Digital Identity Management Framework

Student: LUM Chun Ho

Objective: The traditional way of identity verification require to access through human's certified documents such as ID cards, passports, birth certificates, which usually disclose personal information that beyond necessaryTo make use of Blockchain technology and create a system that enable an easy uploading and mangement of identity data for users and to avoid unncessary information disclosure to third-party.

Fast and Stable Photorealistic Video Style Transfer

Student: NG Tsz Lok

Objective: Currently, Style Transfer is one of the hot topics in Deep Learning domain. However, most of them are focusing on Artistic Style. There is a lack of Photorealistic Style Transfer, especially on Video aspect. By using a feature transformation based photorealistic image stylization method, Fast and Stable Photorealistic Video Style Transfer aims to achieve photorealistic video stylization within a short time.

Reminiscence Therapy VR for Dementia 

Student: TAI Sai Kin

Objective: Based on Hong Kong background and culture, this project makes use of the VR technology to solve the market demand of increasing no. of dementia elderly in HK, reduce the resources needed in reminiscence therapy, problems of gathering materials & reproduce the early living environment.

Smart Library Management System with SaaS Model and RF-based Indoor Localization Technology

Student: TSIM Tat Yau

Objective: Many traditional libraries still rely on "barcode" for books borrowing/ returning and inventory mangement which are extremely time-consuming and easily caused management problems such as security concern, installability and accuracy.

GAN-Aided Drum Music Composition

Student: WEI Yiran

Objective: This project involves human preference, reserve human's ideas as everyone has creativity motivation. it achieved great results in music generation field, not a type of product that can be used by the public. it also shows the potential for applying in-painting problem techniques in user input based music generation.