»Ukiyo-AI« is a visual exploration of generating Japanese Ukiyo-e inspired still-life scenes with the help of machine learning tools. With guide images and Contrastive Language–Image Pre-training used as inputs, we can drive a Guided Diffusion model that creates a stunning assortment of dream-like scenes. It is a mesmerizing experience to see the variety in compositions and materiality generated by ML, giving us a glance at the limitless creativity available within the ML black box.
Concept, Art Direction
Design, Animation, Production
Thanks to the ML community:
Open.ai, Katherine Crowson, Daniel Russel, et al.
»I leave my brush in the East,
And set forth on my journey.
I shall see the famous places in the Western Land.«