Effect of Duration and Delay on the Identifiability of VR Motion

TL;DR

This study examines how the length of training data and the delay between training and testing influence the ability to identify users based on their VR motion data, highlighting the importance of controlling these factors in privacy assessments.

Contribution

It provides empirical insights into how training data duration and train-test delay impact user re-identification accuracy in VR motion data.

Findings

High accuracy with minimal train-test delay

Training data duration affects identifiability

Train-test delay should be carefully controlled

Abstract

Social virtual reality is an emerging medium of communication. In this medium, a user's avatar (virtual representation) is controlled by the tracked motion of the user's headset and hand controllers. This tracked motion is a rich data stream that can leak characteristics of the user or can be effectively matched to previously-identified data to identify a user. To better understand the boundaries of motion data identifiability, we investigate how varying training data duration and train-test delay affects the accuracy at which a machine learning model can correctly classify user motion in a supervised learning task simulating re-identification. The dataset we use has a unique combination of a large number of participants, long duration per session, large number of sessions, and a long time span over which sessions were conducted. We find that training data duration and train-test delay affect identifiability; that minimal train-test delay leads to very high accuracy; and that train-test delay should be controlled in future experiments.