Using Machine Learning to Identify Lightning in Paintings

Abstract

Benjamin Franklin has been closely associated with lightning ever since his famous investigation into the nature of lightning by flying a kite during a thunderstorm. From 1776 until 1785 he was the first American diplomat in Paris, France. He became a cult figure in France and his portrait was reproduced on everything from coins to snuffboxes. Since Franklin is so intimately connected to the idea of lightning, this study sought to determine whether there was an increase in the number of depictions of lightning in paintings in France around his time there as a diplomat. We used machine learning to identify lightning in paintings from the Louvre museum's database. Although there did not seem to be an increase in paintings of lightning around the time that Franklin was in Paris, the study is nonetheless an interesting application of machine learning to art works.