Wellcome Collection

Two interactive exhibits, a real-time average face of visitors and creating a visual identity using biometric data. Created for Wellcome Collection at Ico Design.

Jul 2007

The Wellcome Collection is a major new public venue for London. Ico Design were commissioned to design and produce interactive installations that would intrigue visitors and cause users to question preconceived ideas of biomedical data and what it means to be human.

The brief for the first permanent exhibit was to create a biometric identity, a visual representation of a visitor generated by their bodies data.  The second was to give an impression of the average Wellcome visitor, so we created what is believed to be the worlds first fully automated ‘average face generator’.

Biometric Identity

The visitor uses the touchscreen to go through a step by step process of building up their visual identity. First their pulse is taken, then their fingerprint scanned, height measured, age entered and iris photographed. At each stage the visitor can see how their data is constructing the logo graphic. Various attributes are designed based on the data, such as shape type, line thickness, fingerprint rings and colours from the iris scan. Visitors can then compare their identity to others created previously.  Behind the interactive is a large attractor screen that shows a visualization of various biometric identities, how they were constructed and how they group together.

Average Face

The visitor takes a photo of themselves and the software automatically estimates 86 points of their facial features. They are then shown people who statistically look like them. The user is shown how their facial proportions compare to the average face, followed by a live morph of their photo into an average shape. Then the average face is generated from the last 50 people who used the exhibit.

The user answers a series of questions such as age, sex, eating habits and health. With this data collected for each visitor, it is then possible to see the average face of different groups, i.e. The average female who smokes, or the average over 40 year old who is vegetarian.

My role
Technical lead on the project. This included writing all software, specifying hardware, advising on mounting, circuit design, specifying database development and overseeing attractor screen development, install and robust testing.

Technical details

Working closely with the development of the Betaface technology for face tracking and morphing, through features ideas, integration with Director as an Xtra, sample Director code and robustness testing.

The interactive software was programmed in Adobe Director. A PointGrey Dragonfly camera was used. Arduino was used for height sensor inputs and to fade the LED behind the white corian surface. 

The exhibit used two connected computers, one to run the interactive and the other for the attractor screen animation. Images were automatically formatted for online publishing and for display on attractor screen.