It takes 16,000 computers to recognise a cat

IT TAKES 16,000 computers working as a brain to tell if a cat is a cat, according to a study.

The results of the study will not be released formally until later this week, so for now we must be content with a sneak peek at the New York Times. There we learn that an experiment to create a computer brain carried out in Google's X laboratory created a machine with the ability to recognise a cat.

The neural network made up of 16,000 processors was let loose on the internet, reports the NYT, and given the opportunity to learn. What it learned is what a cat looks like.

There is more to it than that, of course. After all, even a dog knows what a cat looks like, and in their abstract the researchers said that the work will have benefits for face recognition systems.

"Contrary to what appears to be a widely-held intuition, our experimental results reveal that it is possible to train a face detector without having to label images as containing a face or not. Control experiments show that this feature detector is robust not only to translation but also to scaling and out-of-plane rotation," it said.

"We also find that the same network is sensitive to other high-level concepts such as cat faces and human bodies."

The results of the study will be presented in Edinburgh this week. µ