For this looking outwards I mainly looked at papers that related to extracting patterns from 2d visuals.

Pattern Recognition Using Genetic Algorithms
In this paper the author recounts how his approach to designing “creatures” in a genetic algorithm and how they perform at recognizing patterns in 2d visuals. I thought this was interesting because the author emphasizes drawing from existing visual systems in animals and refers to that in his design. One problem with this paper is it reads like a journal entry about ideas that may potentially be more fully explained later but are not quite fleshed out yet; given this it was hard to follow some of the paragraphs that trail off into different possible explanations for the observed results.

A Language for representing and extracting 3D semantics from Paper-Based Sketches
I liked this paper a lot more because the application of the research was not unclear; I think its really interesting to think of pattern recognition in terms of recognizing parts of a 3D geometry and not just the repetition of 2d patterns like the previous paper. This paper also appealed to me because I find the idea of paper-based programming and languages that are not linear (but spatially organized) super fascinating. The goal of the paper is to allow sketching in conjunction with annotation that defines operations (i.e. “extrude”, “sweep”, “revolve”) to result in 3d forms.