[D] Artificial Nature
During the research, a StyleGAN2 model was trained with datasets of three different nature areas and a total of 1.500 images were generated (500 images per nature area). After generating all the images, they were sorted out in two different categories bearing in mind my perception of reality: conceivable or unconceivable.
The project consists of two parts: a small installation of three monitors, which showcases two videos per nature area and aims to show the differences between the conceivable and unconceivable images, and a catalogue, which shows the total amount of generated images.