New research from Stanford indicates that VR users could be identified using only 5 minutes of motion data

New research from Stanford indicates that VR users could be identified using only 5 minutes of motion data

Researchers at Stanford University have shown that they can accurately identify people after only a five-minute session in a standard VR headset.

According to the German tech website, MIXED, the researchers developed a system that can identify individuals in “typical VR viewing circumstances, with no specially designed identifying task.”

How they did it

A pool of 511 users were given a standard HTC Vive headset and wand controllers, they then watched five 20 second clips from a random set of 360-degree videos and answered questionnaires in VR. The videos were designed to see how users would react and move in a VR setting.

All of the recorded non-verbal data was plugged into a set of three machine learning algorithms. This created a profile for each of the users that measured their height, posture, head rotation speed, distance from VR content, position of controllers at rest, and how they move. The user could then be identified with 95% accuracy after only 5 minutes of non-verbal tracking data.

What does this mean?

This has huge implications for privacy. The research paper notes that “In both the privacy policy of Oculus and HTC, makers of two of the most popular VR headsets in 2020, the companies are permitted to share any de-identified data. If the tracking data is shared according to rules for de-identified data, then regardless of what is promised in principle, in practice taking one’s name off a dataset accomplishes very little.”

So, despite your explicitly personal user data being private, the other, technically less identifiable, data could provide companies with all the information they need to accurately identify each person and their habits with a shocking degree of detail.

The research paper ends with a warning and a request, “With the rise of virtual reality, body tracking data has never been more accurate and more plentiful. There are many good uses of this tracking data, but it can also be abused. This work suggests that tracking data during an everyday VR experience is an effective identifier even in large samples. We encourage the research community to explore methods to protect VR tracking data.”


Andrew Boggs

Andrew is a Northern Ireland based journalist with a passion for video games. His latest hobby is watching people speedrun Super Mario 64 and realising how bad he is at platformers.

Andrew Boggs

Related posts