Research Projects

Key Research

• Storage and management of large and varied datasets
• Capturing, digitizing and processing data
• ​Data analytics and visualization​

Key Research

• ​Out-of-core massive building and environmental data management for VR
• Map representation generation and dynamic path planning
• Mobile edge computing
• In collaboration with Beijing University of Posts and Telecommunications

Mixed Reality Fire Safety application which features an interactive emergency scenario where the fire breaks out within the work area while outlining the necessary steps required to escape to safety. The application was developed to be used on Microsoft Hololens MR glass.​​​

We aim to improve the cognitive capability of children and adults through two interactive brain training games, i.e. Smart Brain Trainer and Smart Cognitive Assessor. Smart Brain Trainer contains ten games with different gameplay aiming to train various cognitive aspects. Four of the games use vocal word detection, i.e. players are required to speak out the correct word answer based on the questions presented in screen (e.g. they need to say ‘green’ when a green arrow is shown in the screen). Smart Cognitive Assessor contains eight test modules for assessing various cognitive aspects. Originally, these two games are conducted manually with a staff/trainer supervision. In this project, we digitize the two games as the digital version is scalable and more efficient compared to the manual version. In the manual version, one staff/trainer is always required to evaluate/mentor one student. By using the digital version, each student can install the application and do the assessment/training from his/her own workstation without requiring one personal mentor. We have also developed a cloud platform for supporting the games. User data such as account, license, game progress, and other data are store in the cloud.

This project is done in a collaboration with Singapore Brain Development Centre (SBDC), a training/rehabilitation centre for kids and adults with attention focus problem, ADHD, Autism, Dyslexia, and more.​​​​​​​​​

In the traditional ‘laser pointer’ method of diagnosing neck pain patients, the doctor places a laser pointer on the patient’s head and asks the patient move their neck based on the doctor’s command. The doctor records the pointer movement on a white board in front of the patient using a camera and analyses the trajectory. This approach apparently is not effective enough since it involves a lot of manual work.

Hence, we develop a new method to study the effectiveness of the virtual reality devices for studying the mobility of neck pain patients. We digitize the traditional ‘laser pointer’ method to the contemporary virtual reality (VR) method. In this project, the implementation is divided into two applications, the doctor-side PC application and the patient-side Gear VR application. Instructions are sent from the doctor-side application to the patient-side application and patient's neck movement data are sent from the patient-side application the to doctor-side application for every frame. In patient's view, they can see doctor's instruction inside the Gear VR device. The movement data of the patient are visualized in the doctor-side application in real-time. We use Samsung Gear VR in this project. 

An efficient resource scheduling scheme for out-of-core dynamic streaming of a 3D scene. The entire scene is stored in the cloud and relevant scene data are streamed to a client Android mobile device in real time based on a user-selected path in the 3D scene. This is useful for standalone VR devices such as Oculus Go which has limited resources.