Thesis Project Form

Hand?tracking in VR introduces latency, jitter, and incomplete joint reconstruction, which distort natural grasp biomechanics. These inaccuracies limit the realism of VR?based medical training, rehabilitation, and fine?motor skill assessment. A quantitative understanding of how VR alters reach?to?grasp movements can support more accurate interaction models and improve training transfer to real?world tasks.

The objective of this thesis is to quantify biomechanical differences between real-world and VR reach-to-grasp movements under standardized task conditions. Controlled grasping tasks involving objects of varying size and orientation will be designed to elicit reproducible motor patterns. Motion-capture data will serve as a biomechanical reference and will be compared to VR hand-tracking data collected within a Unity-based environment. Key kinematic parameters such as grip aperture, movement time, peak velocity, trajectory smoothness, and joint coordination will be analyzed to identify systematic deviations introduced by VR tracking. Based on these findings, a biomechanical correction or remapping model will be developed to compensate for tracking-induced distortions. The correction approach will be implemented in Unity (C#) and evaluated through controlled user experiments comparing uncorrected and corrected interaction. Performance improvements will be assessed using quantitative kinematic metrics and statistical analysis.

The application will be developed using Unity 3D (scripting in C#). Optimization of the application will be performed in a loop of development and assessment. Participation in the definition of an experimental protocol. Participation in experimental sessions. Analysis of experimental data.