The Internet is strewn with photos and videos of cats, but how many computers would it take to recognise our feline friends? According to new research, about 16,000.
As detailed by the New York Times, researchers at Google have been working on a way to map the human brain. Using 10 million images pulled from YouTube videos and a connection of 16,000 processors, the team was able to teach the machines to recognise cats.
While that might seem silly, the project is noteworthy because researchers never prompted the computers to be on the lookout for cat faces. Over time, the machines just recognised the animals.
"It performed far better than any previous effort by roughly doubling its accuracy in recognizing objects in a challenging list of 20,000 distinct items," the Times said.
The large amount of data and computing power also make the project unique.
The study - conducted by Stanford University computer scientist Andrew Y. Ng and Google fellow Jeff Dean - is part of Google X, a lab within the search giant focussed on futuristic efforts like the interactive glasses experiment Project Glass. Going forward, however, the "cat project" will move to Google's core search business and could be used for things like image search improvements, speech recognition and machine language translation, the Times said.
Ideally, the work could lead to a full, digital map of the human visual cortex. Researchers cautioned, however, that much work remains to be done.
"It'd be fantastic if it turns out that all we need to do is take current algorithms and run them bigger, but my gut feeling is that we still don't quite have the right algorithm yet," Ng told the Times.