title. DANCERX

date. 2021

city. Istanbul

size. AI generated images

Concept & Theme

While researching with AI, I came up with an idea of combining human dancers’ own way of movement and AI’s perspective on this movement. This creates a new form of dancer made by combination of two minds whom i call DANCERX which allows us to explore the unlimited creative potential of artificial intelligence. Besides these movement choices, dancer also had a ‘body’ as an image, so I gave AI ‘roots’ as a tool to visualize that movement choices. At the end of this research we see a hybrid dance performance from two minds to connect digital realm with reality.

The Technical Principle

StyleGAN2 trained on dancers improvisations and tree roots dataset. 

I used machine learning system through StyleGAN model to make a data visualizing output. And I made a sound design for that output to emphasize the story behind it. I used Ableton and adobe audition.

I used an online platform called Runway which is powered by machine learning. Runway run machine learning models. It is important to note that models are trained on the data given to them, which means that they incorporate the biases of that data.