Are my hands enough? Using Inverse Kinematics to improve User Identification
Description
This project investigates a new approach to user identification in VR by their movements. Data is collected over 10 sessions from a weekly VR seminar to ensure transferability of results to multiple sessions. A deep learning model is trained to examine whether or to what extent it can identify all 13 participants.
Results show that VR movement data is identifying. Yet, maximum mean identification accuracy at .36 seems to low to be of use in real world applications.