The camera on the slightly creepy arm takes a picture of the pages in the book, the software uses OpenCV to extract faces, and the faces are passed to Google Auto ML Vision comparing the faces to a Waldo model. The result: There’s Waldo.
Which sandwich do people not like the most? The winner: the Cheese and Tomato, if that even counts as an actual sandwich.
After seeing a 1950s physical visualization, I wondered if I could follow a similar process using modern techniques.