UTAR Strategic Fund (UTARSF) 

Title: Project 04: Privacy-preserving Federated Learning Framework for Secure Vehicular Surveillance System 

Duration: 1 January 2025 – 31 December 2027

Project Members:  Goi Bok Min, Yap Wun She, Denis Wong Chee Keong, Lee Wai Kong

Abstract: Smart vehicular surveillance systems collect a lot of information from vehicles, and most of it is closely related to the drivers and passenger’s lifestyle. This creates a concern that such information could leak the privacy of users if there is no proper protection available during the communication. Federated learning (FL) is one of the promising technologies that can be used to protect the privacy of users, in which only the trained model, not the raw data, is shared with the cloud server. However, such an approach is insufficient to prevent the honest-but-curious cloud server from exploiting the collected information for its own interest and benefits. Fully homomorphic encryption (FHE) is a commonly employed technology to allow computation in the encrypted domain so that the cloud server can learn nothing from the users. This allows the users in FL to encrypt their locally trained model to the cloud server, and perform global training without decryption, thus being able to protect the user’s privacy. However, FHE suffers from heavy computational requirements, hindering its practical performance in AI-powered vehicle surveillance systems. In addition, some existing secure vehicular communication relies on the Paillier and Elgamal homomorphic encryption schemes are insecure in the post-quantum era, because a powerful quantum computer can break such cryptosystems in a very short time.

In this project, our team aims to solve the following research problems:

Korean Ministry of Science and ICT International Collaboration Fund

Title: Development of GPU-optimized data protection technology for centralized and federated learning in AI

Duration:  36 months (1 May 2024 - 30 April 2027)

Project Members: Dr. Lee Wai Kong (Lead), Prof. Goi Bok Min ,Dr. Yap Wun She, Dr. Denis Wong Chee Keong (all UTAR),Dr. Ooi Boon Yaik (UTP), Prof. Seong Oun Hwang (Gachon University, Korea). 

Abstract: Federated learning (FL) is an emerging research topic that has attracted a lot of attention recently. FL allows multiple clients to collaborate and learn a global model without sharing their raw data. On one hand, this allows a privacy protection on the data, but on the other hand, it leaks information due to the transmission of sensitive data (i.e., model parameters) over insecure communication channels. These concerns can be addressed using fully homomorphic encryption (FHE). However, FHE is known to have slow execution speed, thus requiring performance optimization before it can be used in practical applications.

In this project, we aim to propose a FHE protected FL system with high performance. This can be achieved through GPUs as accelerators, and some novel techniques to pack/unpack data to allow efficient parallel implementation.

MOHE FRGS Grant

Title: Quantum Image Encryption Utilizing Quantum Image Scrambling Circuit based on q-Deformed Chaotic Maps for BRQI Model

Duration: 1 December 2025 - 30 November 2027

Project Members: Goi Bok Min, Yap Wun She, Denis Wong Chee Keong, Lee Ming Jie

Abstract: This research introduces a quantum image scrambling framework based on bitplane representation (BRQI), which efficiently stores position, bitplane, and color channel data. To improve the randomness of the key space and the dynamics of the system while avoiding the complexities associated with high-dimensional chaos, it uses a q-deformed chaotic map from quantum group theory. The proposed framework improves image security against quantum threats by providing a more dynamic and secure encryption key mechanism.

Funding Amount: RM58200

Duration: 1 July 2017 - 30 Jun 2018

Team: Yap Wun She (Leader), Tee Yee Kai, Lee Wai Kong, Goi Bok Min, Madhavan, Tay Yong Haur and Victor Tan 

Project Title: Data Classification and User Behaviour Prediction

Description: Consists of two machine learning related projects, i.e., predicting user behaviour and document classification. This grant funded three postgraduate students. Research outputs include software prototypes, possibilities of patent and publications. More importantly, this is the first grant received from industry for CCS.

Funding Amount: RM134,000

Click here: Presentation Slides of ICOIN 2018

External Grant by MOHE FRGS

Duration: 15 Aug 2017 -14 Aug 2019

Team: Denis Wong Chee Keong (Leader), Lee Wai Kong, Yap Wun She, Goi Bok Min, Chia Gek Ling, Ang Miin Huey

Title: Code- and Lattice-Based Signature Scheme Defined Over Extra Special p-group

Description: It is our strong belief that the use of extra special p-group algebra is an equivalent if not potentially stronger form than LWE over ring as group algebra is a more general algebraic structure compare to polynomial ring used in LWE over ring. To demonstrate that our scheme has good performance, we will present some experimental results which are based on a software implementation.

Funding Amount: RM58, 200

External Grant by MOHE FRGS

Duration: 1 Aug 2016 - 31 July 2019

Team: Goi Bok Min (Leader), Chang Yoong Choon, Chai Tong Yuen, Lee Sze Wei

Title: Versatile Coding-Independent Authentic Video Sharing Based on Transcoding-Resillient Watermarking

Description: To be updated

Funding Amount: RM99,300

External Grant by MOHE FRGS

Duration: 1 Aug 2016 - 30 Jun 2019

Team:  Chai Tong Yuen (Leader), Goi Bok Min, Tay Yong Haur

Title: Iris Localization, Segmentation and Recognition Framework Under Non-Cooperative and Less Constrained Environment

Description: To be updated

Funding Amount: RM67,200

External Grant by MOSTI Science Fund

Duration: April 2015 - Sept 2017

Team:  Yap Wun She (Leader), Goi Bok Min, Jin Zhe, Tan Syh Yuan, Heng Swee Huay

Title: Design and Development of a New Symmetric Key Generation: Application to Multi-Factor Authentication Solution

Description: This is a multi-institution grant involving UTAR and MMU. This grant funded two master students in the area of biometric security. Both students proposed template protection techniques based on voice and fingerprint biometrics. Research outputs had been published in Pattern Recognition, Journal of Network and Computer Applications, IET Communications, Computers & Security, etc. An additional grant was also secured by Jin Zhe from Nanchang Hangkong University, China.

Click here for presentation slides of Cryptology2016, Invited Talk at Monash 

Internal Grant by UTARRF

Duration: 21 Dec 2020 - 20 Dec 2021

Team: Sim Hong Seng (Leader), Goh Yong Kheng, Chua Sing Yee, Leong Wah June

Title: Stochastic Gradient Descent Algorithm with Multiple Adaptive Learning Rate for Deep Learning

Funding Amount: RM24,000