click on “+” in upper right to expand images
Works in this series use image generating machine learning models (StyleGAN) trained on images of 1,007 historical still life paintings to create the image printed on canvas. A digitally-designed large format vinyl stencil is then applied to the surface before being painted on with oils. When the stencil is peeled off it creates hard edges, and preserves areas of the print.
The detail above shows the hard-edge contrast between paint and reproduction on canvas. A still life image created by artificial intelligence is printed on canvas. Then, a commercially produced vinyl stencil is applied to the surface before being painted thickly with oils. When the stencil is peeled off it leaves behind hard edges, and preserves areas of the print. Paint and reproduction are intentionally disguised into one another in order to make them indistinguishable in some areas.
Animation generated by artificial intelligence software of all generated output images from a dataset of 1,007 historical still life images. This gives a sense of the look and variety of images generated by the StyleGAN.