Bird House (confidence = 6.230%)
Joanne Hastie original
Acrylic on Paper
11″ x 14″
This abstract painting on paper was created as part of my Machine Learning Abstract paintings. I am painting these abstracts by hand while using Python code and TensorFlow software to generate and classify the paintings. It was interesting for me to see how the digital process transferred to painting as often what works on a screen doesn't transfer to paint (and vice versa). After I finished this painting, I used the TensorFlow classifier to suggest what object the abstract painting could be of and use that as a title. Of the 5 objects it suggested, I title the painting “Birdhouse”.
Inception v3 results:
Band Aid (score = 0.27086)
book jacket, dust cover, dust jacket (score = 0.12527)
birdhouse (score = 0.06230)
binder, ring-binder (score = 0.03644)
pinwheel (score = 0.03004)
ABOUT THE PROJECT:
I wrote Python code to collage and generate compositions of hand painted geometric shapes and then use an image classifier to rank the compositions based on my preferences.
Over time, I update the classifier as I hone my eye as to what looks good as a painting versus a digital thumbnail. After it is painted, I use an untrained classifier to title the painting. I get titles such as “Bulletproof Vest”, “Traffic Lights” and “Envelope”, because they are flat paintings (minimal depth) – most paintings are classified as “Envelope”. Can you see the objects in the painting?
This work has been featured in NeurIPS Online Gallery in 2018, at ‘Women in Data Science’ show in Zurich in 2018 and in RUSI Journal 2019.
To learn more about this project click below.