Undergraduate Secures PhD Scholarship Following Research Success in VR Security

3 Sep 2026
Professor Henry Dai (left) with Gengyang Xu (right), whose outstanding undergraduate research in VR security has led to a RAID 2026 publication and a fully funded PhD scholarship at HKUST.

Gengyang Xu, a 2025 Computer Science graduate, has secured a fully funded PhD scholarship following the success of his undergraduate research in virtual reality (VR) security. Under the supervision of Professor Henry Dai, Gengyang collaborated with researchers from Sun Yat-sen University and Lancaster University to develop VR-Themis, a scalable framework for detecting cloned VR applications.

Their paper, “VR-Themis: A Scalable Framework for Virtual Reality Application Clone Detection”, has been accepted for presentation at the 29th International Symposium on Research in Attacks, Intrusions and Defenses (RAID 2026), one of the leading international conferences in computer security. The paper was selected through a highly competitive review process, with an acceptance rate of 16.6%.

Following his graduation from HKBU, Gengyang has begun his PhD studies at the Hong Kong University of Science and Technology (HKUST) with a full scholarship, reflecting the strong research training and innovative environment provided by the Department during his undergraduate education.

A Novel Approach to VR Application Clone Detection

As virtual reality technologies become increasingly popular across gaming, education and entertainment, the expanding VR ecosystem faces growing security challenges, including application cloning and repackaging. Unauthorised copies of VR applications can infringe intellectual property rights, increase maintenance costs for application marketplaces, and expose users to potential security and privacy risks.

Unlike conventional clone-detection approaches that rely mainly on source-code analysis or user-interface similarities, VR-Themis examines three-dimensional scene structures, virtual objects and interaction behaviours within VR applications. The framework employs a two-stage detection process that enables efficient analysis of large numbers of applications while maintaining a high degree of accuracy.

Strong Results and International Recognition

The research team evaluated VR-Themis using a dataset of 4,277 Unity-based VR applications, the largest dataset assembled for VR clone detection to date. The framework identified 307 suspected cloned applications across 977 application pairs with zero false positives, demonstrating its effectiveness in uncovering real-world cases of VR application piracy and repackaging.

“This achievement demonstrates how undergraduate students can make meaningful contributions to cutting-edge research through close faculty mentorship and international collaboration. We hope Gengyang’s success will inspire more students to engage in research and explore opportunities to contribute to advances in science and technology,” Professor Dai shared.