Behaviour Privacy: Non-verbal Threats in Avatar-based VR Systems
Dilshani Kumarapeli, Sungchul Jung, Robert W. Lindeman · 2022 IEEE Conference on Virtual Reality and 3D User Interfaces Abstracts and Workshops (VRW) · 2022
The current move towards highly-expressive, user-controlled avatars in VR creates unprecedented possibilities for realistic collaboration, approaching face-to-face experiences. This precise tracking of VR user behaviours also provides rich data sets for machine learning and deep learning algorithms to use to accurately identify user be-haviours, such as emotional states. However, the availability, type and amount of data also presents threats to user privacy, such as unobservability and undetectability. The current study highlights the privacy threats these avatar systems can present to users.