Bubamara OCR (94/100)
A desktop application with achromatopsic computer vision for interpreting images as texts, with ladybugs as alphabetic intermediary
"Curious what would happen if I could ask a computer to interpret an image as a crude sequence of letters, I wrote a program, BUBAMARA-OCR, that could infer a text based on Latin alphabet from an image using a computer vision algorithm. Ladybirds, chosen for their roles as predictive and fateful messengers, both biologically observed and reimagined by generative adversarial networks, were the intermediary between photo and alphabet — each letter having its own corresponding ladybird. Whatever ladybird — each section of the image looked like most determined the letter that the program would output."
offline, interactive, silent, macos and windows. By Adriana Ramić, 2018.
