This is an experimental project I did at the School for Poetic Computation. I hope to build on this idea more in the future.
We learned about Jason Salavon's image averages in one of the classes. I wanted to apply this methodology to fashion runway images I had scraped from Vogue.com using Python. This is the average of every runway model in their Spring 2019 database generated with Open Frameworks:
Over the last couple years, the fashion industry has treated diversity and inclusion as a trend. Based on this image alone, it looks like the average of fashion hasn't actually changed. I applied the same approach to individual designers. Designers of color like Luar, Pyer Moss, and Telfar tended to have a darker-skinned average. Note that these brands also have unisex shows:
Chromat, a swimwear brand, known for creating garments for all body types also breaks apart from the generic body shape in their averages:
Comme des Garçons by Rei Kawakubo averages light skinned, but her averages create an interesting effect because they're all so different and avant garde.
Most European brands perpetuate the idealized model body and also tend to be less innovative from season to season.
I'm hoping to expand on this idea in a 3D space while also analyzing the upcoming Fall 2019 shows. You can see more work like this at @lalizlapoet on Instagram.