Terror Incognita

The West experiences war through images that glamorises weapons and destruction. Aerial photography, which is usually made at the request of Western governments, provides a unique view of the aftermath of bombarded cities. It abstracts the city, exposing the skeletons of destroyed buildings. I have trained a Generative Adversarial Network (GAN) on thousands of these satellite images, which picks up on repeated visual patterns in the photographs. It uses them to create a new image, producing an average of the dataset. These are then stitched together to form a map of a destroyed city that does not exist. The averaging of the GAN and the repetition of the collage remove us further from the moment of destruction, highlighting the hyperreal nature of images of war.