For xCoAx 2020 I presented an installation that generates a three-dimensional artificial skin. The skin is trained with machine learning on a collection of Tinder profile pictures, scraped from the matches of an account made for the research.
A semantic-segmentation model finds skin regions in images of bodies. After training, any combination of 2D shapes in the same schema lets the system grow images of skin.
labor5020.ch/convolutedAlterity
Team
Andrés Villa Torres.
Technologies
Python, PyTorch, computer vision, web, Colab, Java, three.js, machine learning.
Partners
xCoAx, Pro Helvetia, MUMUTH, Forum Stadtpark, IEM.
Image placeholder — cover
Image placeholder — process