165 Faces from the Yale Face Database
Using the freely available facial recognition database of Yale faces, I did a remix of the 165 images it contained. It was made using a Java application I wrote and tested earlier here
450 Faces from the Caltech Database
Update: Face detection applied to the scatter-average images
The Living Portrait Project is a participatory networked photo installation that was exhibited in 2003 at the d.u.m.b.o. arts festival in Brooklyn (archive here). As a way to get people involved in the exhibit and work collectively on the outcome of the project, the portraits were all averaged together to create a universal portrait of everyone together. I also averaged together groups of 50, as the portrait began to blur beyond recognition at that point.
As an upgrade to this old project, I’m looking into how I can reuse the database of images to kickstart another project, a real-time portrait processor installation. This idea might be carried out with a webcam that webcam and distance sensor that allows a user to mix themselves into the “universal” image through their interaction with the other images within it.
Above is a test run for how the images look when they are aligned and un-aligned. Both images are random mosaics of the individual pixels within each image. The aligned image uses a database of 800 pre-aligned images, and the image on the left is same process with the raw photographs.
The previous Living Portrait Project
Subscribe to RSS