Demystifying biometric data use

22nd May 2019
Facial recognition, an AI-driven process that seems like a “black box” at first, can be easily explained with live overlays, removing some of the uncertainty around it.
Product Context

Qlue participated in the TTC Labs Data Innovation Program, which was part of the first season of Startup Station Singapore in partnership with IMDA.

Qlue is looking to use facial recognition in a workplace HR context to monitor workplace attendance and office security and safety (i.e. by preventing unauthorised visitors). Given this context, they need to explain to people how this technology works, and what they are using it for.

In order to provide the service, Qlue’s app is powered by a training dataset of pre-existing photographs (the subjects of which have consented to being included in this dataset), and requires participating staff to photograph themselves to enrol in the application.

Problem & Opportunity

While people are now pretty familiar with using biometric data on an OS level — to unlock access to their phone for example — the collection of such information by apps often makes people feel uneasy.

How might we... people get more clarity and comfort with the technology behind, and uses of, facial recognition?

Team Qlue discuss approaches to their HMW statement

Design Features
Visualising the mathematics behind facial recognition

A simple, dynamic visualisation of the points collected and analysed by the facial recognition algorithm, coupled with some explanatory text, could help people feel more at ease with how the technology works, and understand that facial recognition is ultimately just a series of measurements.

The aspiration here is to demystify the technology powering the system by exposing some of its logic in a way that is understandable and approachable, and self-enrollment using someone’s own device is a good opportunity to demonstrate how the technology works in a controlled environment.

Design Features
Clearly explaining what data is used, and why

Before someone self-enrolls in the program, the app clearly explains exactly what personal data it uses, and why. A clear link is drawn between the data collected and its relevance in a workplace context.

Additionally, a short video explanation is offered, for people who are interested in learning more about the technology before getting started.

Next steps

As facial recognition becomes a bigger part of more day-to-day interactions, people could benefit from better understanding how the technology behind it functions.

How might we build on Qlue’s ideas to help educate a broader public on how facial recognition technology works?