Uncanny Valleys: Generative landscape

See: Main video with project details

We showcase a new approach of generating 3D landscape models using GANs and U-Net deep learning models. Interpolating the latent space we generate an animated sequence of 3D landscapes.

​

^ Sneak peek ^

If you are in Zurich during the ArtSci 2019 exhibition between 29.4. and 10.5., come visit us at the CHN building of the ETH Zurich campus and also check out the physical, 3D printed representation of this project.

​

Team: Vit Ruzicka, Alexander Nikolas Walzer, Nizar Taha

Open source code: our AerialNets and NVIDIA’s Progressive GAN

​

#MachineLearningForArt, #DeepLearning, #GenerativeArt, #GenerativeLandscapes