The human brain is good at sorting through large amounts of data, finding patterns in that data and making sense of it. It does this thanks to a series of brain cells, called neural networks. Geoffrey Hinton and graduate students, Alex Krizhevsky and Ilya Sutskever, at the University of Toronto’s Machine Learning laboratory have developed learning algorithms for computers that are inspired by these networks.
With equipment funded by the Canada Foundation for Innovation, Hinton’s team studies how collections of neural networks can learn by changing their interactions with one another. This research led them to develop machine learning models in 2012 that dramatically improved a computer’s ability to recognize speech and identify objects in images.
The team’s spin-off company , DNNresearch, was acquired by Google in March 2013 and the tech giant is now using the company’s approach to recognize objects in untagged images. Users can upload their personal photo collections to Google+ and search for photos containing any of 1,100 different types of objects. This technology has also been applied to Google’s Android phone.