Scientists have learned to recognize smeared into the image objects and text

The neural network is able to “see” the object in the picture, comparing it with the original.

Researchers from the University of Texas has developed a system of machine learning that is able to identify smeared faces and text.

So far, the system was not able to razmenivat” picture, but it is able to identify the blurred object, comparing it with the original. During testing, the neural network was able to recognize 80-90% of your processed images on YouTube and 50-70% of the images, smeared with the help of photo editors.

But in the case of images processed using the encryption tool P3 (Privacy-Preserving Photo Sharing) accuracy was about 17%.

Related posts:

Who will be the President of the United States The best job in the competition, Siena International Photo Awards. Photo Named the best funny pictures of wild nature. Photo USA: 100 billion for brain research