Continual learning has always been an interesting topic to me. We as humans learn in a dynamic environment that’s constantly changing, but we have no trouble retaining our learnings from past examples and use it on new examples. In contrast, a dee...

Recently I’ve been reading about how to approach representation for 2D shapes. I’m curious since, as an 2D artist, we use shapes to build up characters, objects, scenes, compositions, and they seem to be the building blocks of the resulting image....

In the previous post we’ve developed a texture generating GAN that is able to convert a small patch of noise to a higher resolution texture: Using the same networks, I created an application that can convert a small drawing with color blobs int...

Recently I have been reading about texture generation and interpolation, and I thought it’d be something interesting if we could generate textures with GANs. Ideally, similar to the GauGAN work, we can specify the layout of an image and let the GA...