Presented by:

Ken Kahn

from University of Oxford

The Noisy Polygon project that draws "polygons" with the inputs for size, angle, pen width, and color randomly chosen between a range of values. You rate each generated noisy polygon and then a machine learning program is trained to predict how you will rate new polygons. One application of this predicts ratings for hundreds of images and displays the one the system thinks you'll rate the highest.

5 min
Main Room
Snap!shot 2021
Lightning Talk