An exhibition about the use of unconstrained and non-consensual data sources collected "in the wild" in Artificial Intelligence systems
The exhibition Face First: Researchers Gone Wild looks at the use of unconstrained and non-consensual data sources in Artificial Intelligence systems. Beginning in 2007 with the "Labeled Faces in the Wild" dataset and continuing throughout 2020, the practice of collecting images "in the wild" has become normalized, but is still largely unregulated and often problematic. Today, dozens of image training datasets with millions of images and identities feed into industrial facial recognition systems with little oversight. The works presented in Researchers Gone Wild showcase discoveries made during the multi-year research project MegaPixels that investigates image training datasets created "in the wild".
Read more on Eigen Art + Lab