The reversed face seems similar to the normally positioned face, until you click on it (Try it!).
Hillary Clinton has her mouth and eyes upside-down! Since we normally do not see faces reversed ,we are fooled, which is not the case when we see the face in its normal position. In the latter case the inconsistencies are easily recognized.
Even though we do not exactly know the brain mechanisms underlying face recognition, scientists have demonstrated that we have a specific brain area devoted to this task.

2. How does our brain respond to faces?
Certain neurons respond only to faces. They do not respond to pictures of houses or other objects. Some neurons stll respond to faces without eyes, or cartoon pictures of faces. The brain area involved in face recognition is the fusiform face area (FFA). Some neurons in this area respond to frontal faces, while others respond only to profiles. Other neurons react specifically for certain facial expressions, like sadness.

3. Where in the brain is facial information processed?
The fusiform face area (FFA) processes facial information. It is situated in the rear of the brain in the temporal lobe.

4. How do we know that we have a special brain area for recognizing faces?
Specialized research has revealed information about face recognition. From neuropsychological studies, it is known that certain patients have difficulties in recognizing faces, due to a disturbance in the FFA.
This particular syndrome is known as prosopagnosia, which is derived from the Greek words prosopon (face) and agnosis (without knowledge). These patients still know that they are looking at a face, but they cannot identify the person, even family and friends that they encounter frequently. These patients often use other cues to recognize people, like the voice or particular body movements. Some of these patients even cannot recognize themselves in a mirror.
Neurophysiologists have found specific brain areas involved in face recognition. In one experiment, instructed subjects looked at pictures of faces while their brain activity was simultaneously recorded using functional MRI. Results from these studies demonstrate that a particular brain region is involved in face recognition.

5. Why is recognition of faces important?
First, recognition of a face helps to identify people, their sex, age, race etc. This information is important for social behavior. Furthermore, it can be used to recognize other people’s emotions. In verbal communication, not only is speech crucial, but also the movement of facial features, like the mouth and lips are important cues. Not only do deaf people use this information, but also everyone can use these cues, especially in noisy environments.

6. When do babies start recognizing faces?
At three months, babies can easily recognize faces of known people. Even hours after birth, they can already see patterns, although their visual acuity is not that sharp. Patterns with an oval shape and two dots at the correct position are interesting for these young babies.

7. Is the recognition of faces learned or innate?
Face recognition is innate, but requires training. In one study, subjects were taught to recognize fantasy forms. After training, their performance matched that of their performance in face recognition. Forms and objects, therefore faces, that we encounter regularly, and those that are important to us, are easy to memorize. For Western people it is often hard to distinguish the differences in Asian faces. After staying in Asia, a person’s ability recognition improves greatly because they learn to look at facial details.

8. Why can't a computer recognize objects or faces as easily as we do?
Human visual perception is not just a matter of the retina analyzing light, dark, or colors. To understand what we see, cognitive features are very important. Among other things, we use the meaning of an object, the characteristics of its environment along with our memories and expectations to interpret the scene. In a way, to perceive, you need intelligence. Even though a computer can win a game of difficult chess, it is not intelligent.
For example, a large, distant square projects the same image on the retina as a smaller, nearby square. Even a tilted trapezoid could create the same image. Humans normally observe these objects in the correct way without thinking, because they can consider the entire context cues that are available. However, for computers this task is difficult. Similar problems occur when an objects are partly hidden behind each other. Additionally, the reason for differences in lightness can be unclear, since different object colors and uneven illumination could account for changes in shade. A white house, of which one wall is standing in the shadows, is still a white house to us, not a gray-and-white one. When we are presented with the same image as a computer receives, we perform much better in understanding what we perceive, thanks to our ability to use our experience.

9. What is FindMyDouble?
FindMyDouble is a website on which anyone can find his or her double. It's definitely not perfect, and is more an entertainment site than a scientific one. Nevertheless, the techniques used to analyse and compare the images are based on on a biologically plausible theory of how the brain processes information about objects and faces. This theory is called "the Scale Space Theory" as proposed by Jan Koenderink. Bart ter Haar Romeny recently published a book on the Scale Space Theory and its applications.

10. Who made FindMyDouble?
FindMyDouble was originally created (2000) by three students from the Delft University of Technology (Rikkert Nachtegaal, Robert-Jan de Pauw and Dennis Hendriksen) and two neurobiologists (Richard van Wezel and Bart Borghuis). This first version was hosted by Vergelijk.nl. For more information, read the Ublad article, written by Rinze Benedictus (in Dutch). Or, read the official report (English, pdf, 500 kB) about FindMyDouble for the Universiteit Utrecht and the Delft University of Technology.
In 2004, the site and the software have been revised and translated to English. The analysis and comparison is now done using the software of two students of the Technische Universiteit Eindhoven. Sander Maas has created the analysis algorithm and software, and Martijn van Tienen handled the comparison algorithm and software (click their names for the reports, both about 600 kB).
Part of this work has been sponsored by NWO.

11. How can I get the best results?
- Take a large, sharp picture with your face in the center (both ears visible)
- Draw the rectangle as close as possible around your face
- Take an image with high contrast (vague, dark webcam shots give poor outcomes)
- Only use pictures with just one person on it (no group shots)

12. I want my image to be removed!
That is possible; you have to use the link that has been sent to you after inserting your picture. When you have logged on, there is a button 'Remove me' below your personal data.
If you have lost this email we can re-send it. Fill in your email-address and you will receive an email with a new link.

13. I have another question.
Mail it and we will have a look. This can take a few days/weeks so do not be too impatient.