6/25 — Week 6 — Image/Face averaging/compositing (Nancy Burson/Jason Salavon)
Opened this issue · 0 comments
Reading recap
Art and Technology, a Panacea that failed
Artist: Nancy Burson
- Artificially aging, based on general morphing technique: break image into triangles,and crossfade/manipulate the triangles in some way (e.g. morph your face into someone else's face)
- Really inspired by participatory art > build machine to age people, "warping" "morphing"
a system for seeing words/phrases in 3 diff directions simultaneously:
"patent pending" thinks about patents a lot
problematic aspect of her work:
"its artificiality revealsnothing about how people live their lives"
in order to create these images it requires stereotyping
eugenecists did composites before, "average criminal face". dark side to face averaging
related to ML biases today
Jason Salavon
Data portraits, image averaging/stats to tell diff stories
Data has an essence that always come thru no matter how much you transform it. It has a core
Past friend/student work
Interactive work, averaging all the photos of people who go up to it
https://www.alifibrahim.com/copy-of-you-are-not-who-you-think-i
Pareidolia (humans seeing faces in stuff)
Take things that look like faces and average them (run thru face detection algo and align and average) you create this ghost
Research
https://paper.dropbox.com/doc/Nancy-Burson--A2nbdC6UErtLUcgN72D4a4YmAg-0Y9tEB5HBSrDjXto4G9G8
https://paper.dropbox.com/doc/Jason-Salavon-TCojb9CiFu1v8eVZU2nxh
Technical notes
- face detection
- averaging/normalize/rotate/scale/align (e.g. align so eyes are always in the same point)