Performed in Zorlu PSM, Istanbul. 16th of September, 2022.
Below you can find the techniques and data used in the Tree of Life audio/visual performance in collaboration with Fazıl Say, Jamal Aliyev and Zorlu PSM.
At the same time, we share the StyleGAN-2 model, which has been trained with 65 million nature and tree images, as open source.
With the latent gan browser, you can see how synthetically produced images are created.
Optical flow pipeline diagram of synthetic images
60 million classified and cleared visual umap algorithms are positioned in a 3-dimensional space.
If you want to view the full video in high resolution, you can visit the video below.
The weights of the model trained in StyleGAN-2 can be downloaded from the link
Some examples of sound analyzes made for a small part of the works are shown below.
Performed in Zorlu PSM, Istanbul. 16th of September, 2022.