210 days ago,4425 views
Presented by: Steve Crow (@cr0wst)
Every human face has its own set of unique features. Isolating these features and determining what goes into making a face, isn't a trivial task. What does it take to make a computer recognize a face?
In this talk he will start by explaining the concept of the Singular Value Decomposition and its use in machine learning. He'll then demonstrate how it can aid in isolating facial features from images of people. Lastly, he'll show an example on how one might construct a facial recognition system using these techniques.
Steve is a self-proclaimed Mathlete, and King of Snark. He is also a lover of Greyhounds, twisty puzzles, and European Board Games. When not talking math to non-math people, and Java to non-Java people, he can be found sipping coffee and hacking on code.