What we work on

VR and XR are moving from entertainment into training, health, and collaborative work. The minute an immersive system becomes safety- or business-critical, the fidelity of its sensing, pose, motion, scene geometry, turns into a security property. A digital twin that drifts a few centimeters off the ground is not just an animation glitch; it’s a trust problem.

The lab works on 3D pose estimation and ground-relative positioning that produces faithful, authenticatable digital twins of human motion. Our methods plug into off-the-shelf pose frameworks and recalibrate the character’s position against the real-world floor plane, so virtual characters interact with their environment realistically.

Where this is going

Beyond visualization fidelity, we’re interested in trustworthy XR sensing more broadly: sensor authentication for VR controllers, attack-resistant SLAM, and verifiable provenance for digital-twin recordings.

Publications in this area

  • ACTAUTH: Interaction-Triggered, Continuous User Authentication for Controller-Free Virtual Reality

    Haoran Li, Omar Achkar, George Zouridakis, Xin Fu, Kyu In Lee

    Accepted in IEEE International Conference on Cyber Security and Resilience (CSR) · 2026

← All research