• DNNs "learn" to recognize images by being exposed to many of those of the same type—they use learning algorithms that spot commonalities between parts of information in the images to map out different aspects of different objects.
  • The researchers hooked a well known and respected DNN called AlexNet to a system that also included algorithms developed to evolve pictures using basic elements. The result was mostly images resembling static, which the computer was certain they were of things (like lions, for example). This is due to the fact that all the parts the DNN is looking for are there, but not in a logical way.
  • The importance of this discovery resides in the fact that DNNs are used in various applications, such as by cars that drive themselves.

Share This Article