You Need to Look at the Eyes to Understand Fake Images

In the age of deep fakes and realistic fake images, a new way of knowing whether the image in front of us is real or fake has been discovered: looking at the pupils of the eyes.

In almost every period of human history, some kind of forgery has. Today, however, deep fakes and fake images that look realistic stands out. It’s getting harder and harder to spot these fake images. Scientists, on the other hand, have an advice for understanding fake images.

Thanks to the method developed by scientists, we can see the faces we look at. find out if it’s fake or real It may be possible. To understand the difference between the images created by artificial intelligence and real people, we need to look at the eyes.

The truth is hidden in the eyes:

researcher at the State University of New York Hui Guo According to his study, the secret to understanding fake images lies in the eyes, especially in the shape of the irises. To understand the difference, all we have to do is zoom in on the images.


Feast Your Eyes: Samsung Announces First mini LED Gaming Monitor

When we get closer to the image before us structure of the pupil and iris It allows you to understand whether the image is real or not. While real irises and pupils are round, distortions occur in artificial intelligence images.

In research team articles “Iris are almost round in shape in healthy adults. Compared to real faces, GAN-made images we observed visible marks and inconsistencies in the eye area.“He used the expression. According to the researchers, the reason why artificial intelligence makes this mistake is that artificial intelligence does not fully understand the anatomical structure of the human eye.

It is possible to distinguish fake images:



Do You Say Texting Him When You Blink? The Dream of Controlling the Phone with Mimics Will Come True with Android 12

Scientists use images in the images to test this theory. eye shape After developing a software, he uploaded 1000 real and 1000 fake face photos to the system. Evaluating eye shapes, the program was able to reliably separate images just by looking at this criterion.

According to the researchers, one day, thanks to this study, your fake images It can be used to take precautions to prevent it from being used as if it were real. It is believed that these measures will be very useful, especially on platforms that are open to manipulation such as social media.

Source :