While the applications of machine learning for text and numerical data tend to be the most discussed, there are similarly a large number of applications for images. Many of these tend to be in areas of cutting-edge research with profound implications.

We are quite accustomed today to running a Google search to find some odd snippet of text from a news story that we read three weeks ago and having the first result be just what we wanted. Now, imagine if we could do the same thing for pictures, movies, and GIFs. For example, let's say you saw a GIF six months ago that is perfectly apropos for your current Slack conversation, but all you can remember is that it involved a llama and a man trying to feed him. Right now, you'd have a tough time locating it because most image searches involve hashtags and the text surrounding the image. Over the last few years, however, companies such as Google have made amazing progress in having machines label the contents of images...