To create these images I've used DCGAN Machine Learning to explore the latent space of visual possibilities within a dataset of images all representing a wide array of 'textures'. The technique is basically the same as this project if you'd like to dig a little deeper.
The resulting generated content has been super interesting.
More to come soon!
![](https://uploads-ssl.webflow.com/5980de29cdb97100016dfb6b/5dd86cb3cd7d37695d14f4ab_eigen_1.jpg)
![](https://uploads-ssl.webflow.com/5980de29cdb97100016dfb6b/5dd86ccd967a3bbe1426474f_eigen_3.jpg)
![](https://uploads-ssl.webflow.com/5980de29cdb97100016dfb6b/5dd86ce3578bab00a018088b_eigen_2.jpg)
![](https://uploads-ssl.webflow.com/5980de29cdb97100016dfb6b/5dc88ce66ade635d14fddf62_train_28_0282-Mono%20Comp.png)