Facebooku0026#x27;s Deep Learning project
Mar 21, 2014
Google recently acquired a company called DeepMind Technologies (they say for over $400 million). Google was also very lucky that Ray Kurzweil said yes on becoming their Director of Engineering and machine learning. Today it’s fashionable to talk about deep learning, or machine learning. Once we called it artificial intelligence.
Facebook calls it “Deepface” when they talk about their AI breakthrough in artificial intelligence. Basically what they’ve built is a facial recognition machine that is as good as how humans recognize faces. To be a little bit more precise, it’s not exactly facial recognition (putting a name to a face) but facial verification (it recognizes that two images show the same face).
Facebook recently released a research paper on the project last week.
How Deepface works
DeepFace processes images of faces in two steps. First, it corrects the angle of a face so that the person in the picture faces forward, using a 3-D model of an “average” forward-looking face. The deep-learning part of DeepFace consists of nine layers of simple simulated neurons, with more than 120 million connections between them. To train that network, Facebook’s researchers tapped a tiny slice of data from their company’s hoard of user images—four million photos of faces belonging to almost 4,000 people.
Then the deep learning comes in as a simulated neural network that works out a numerical description of the reoriented face. If DeepFace comes up with similar enough descriptions from two different images, it decides they must show the same face.
So the question is… Is this the end of making AI jokes like this?