Landvættir Vol 1: Palasip Qaqqaa The Landvættir series explores an alternative approach to generative A.I art. In particular, generative adversarial network art. Landvættir is a further exploration of my #Scenism manifesto. Landvættir‘s #virscapes were not simply generated from image scraping; a ‘mass farming’ of images, but arose from individual curated gameplay experiences; each image is a visual expression of a meaningful and curated playspace, the resulting dataset used to generate the ar…

Landvættir Vol 1: Palasip Qaqqaa

The Landvættir series explores an alternative approach to generative A.I art. In particular, generative adversarial network art. Landvættir is a further exploration of my #Scenism manifesto. 

Landvættir‘s #virscapes were not simply generated from image scraping; a ‘mass farming’ of images, but arose from individual curated gameplay experiences; each image is a visual expression of a meaningful and curated playspace, the resulting dataset used to generate the ar...

Landvættir Vol 1: Palasip Qaqqaa The Landvættir series explores an alternative approach to generative A.I art. In particular, generative adversarial network art. Landvættir is a further exploration of my #Scenism manifesto. Landvættir‘s #virscapes were not simply generated from image scraping; a ‘mass farming’ of images, but arose from individual curated gameplay experiences; each image is […]

Skulls Third experiment with GAN. Trained around 50 images of front and side skulls with different materials for this one. The machine couldn’t draw an exact replica of those real skulls in the amount of training I gave it. It probably needed more time as some material was very detailed (wood, glitter, marble — patterns) but I loved the abstract results, so I stopped the training! Yet, again like oil paintings, that this was my fit. . . . . . . . . . . . . . . . . . #gans #machinelearning #ai…

Skulls 
Third experiment with GAN.

Trained around 50 images of front and side skulls with different materials for this one. 
The machine couldn’t draw an exact replica of those real skulls in the amount of training I gave it. It probably needed more time as some material was very detailed (wood, glitter, marble — patterns) but I loved the abstract results, so I stopped the training! Yet, again like oil paintings, that this was my fit.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
#gans #machinelearning #ai...

Skulls 💀 Third experiment with GAN. Trained around 50 images of front and side skulls with different materials for this one. The machine couldn’t draw an exact replica of those real skulls in the amount of training I gave it. It probably needed more time as some material was very detailed (wood, glitter, marble — […]