Yes, you did read that correctly.
Hidden away in their secret X Lab where work on futuristic projects such as artificial intelligence and robotics is carried out behind closed doors, a recent project focused on a topic slightly closer to home for many of us – searching cat videos on YouTube.
Google engineers have recreated the human brain using a neural network of 1,000 computers which boasts the power of over 16,000 processors.
In order to put it to the test, however, engineers tasked the ‘brain’ with completing a popular human time-sink for itself by searching for cat videos on YouTube.
Throughout this test, the ‘brain’ was shown ten million images from various videos on the video sharing site and had to identify which of these images were of cats. Impressively, the experiment was successful and not only was the brain capable of identifying our feline counterparts, but did so with no guidance or supervision.
Whilst amusing regardless, there was in fact a distinct purpose to the tests which was to discover whether computers are capable of learning facial recognition without labeled images or direction. In addition to cats, the ‘brain’ was capable of developing concepts for the identification of human body parts.
Researchers working on the project claim there was 15.8 per cent level of accuracy in recognizing 20,000 object categories – a 70 per cent increase over previous tests.
Want to receive up-to-the-minute tech news straight to your inbox? Then click here to sign up for the completely free PCR Daily Digest and Newsflash email services. You can also follow PCR on Twitter and Facebook.