Google and Stanford University link 1,000 computers, creating a network of nerves has over 1 billion connections, simulating an artificial brain capable of learning the basics.
Andrew Ng showcased the ability to identify artificial brain of cats-photo: Extremetech
Dr. Jeff Dean, head of the Group of experts from Google's secret lab, along with two other associates came from University of Standford's Andrew Ng and Le Quoc created a system that simulates the brains of babies and have the ability to perceive.
With 16,000 core processing in 1,000 computers linked together, the computer system simulation brain was able to automatically recognize the people and species of cats with high precision.
"We were not told it (computer systems) in the process of training a cat. Basically, it self aware of the concept of a cat "-Dr. Jeff Dean, who heads the project, said with the New Times.
The previous face recognition are usually programmed to identify traits, but Google's artificial brain is not-photo: Extremetech.
According to the technology of The Verge, the Google engineer did not load any characteristics of the cat on the artificial brain can be identified as to how to do it before. Instead, this system to "see" a huge amount of video on YouTube related to cats. Then, an artificial brain began to detect the familiar objects and mapping images of that object.
Over time, an artificial brain would form the preliminary concept of the cat through the data in shape, size, colors ...
According to the newspaper, the New York Times, in the near future artificial brains will be applied for the Google image search. Google said it will also improve the algorithm for artificial brain system can search images, voice recognition and language translation.
No comments:
Post a Comment