News (June 12, 2010)
A special on the World Cup 2010 Football/Soccer: the average player of each national team! Here!.
News (June 9, 2010)
A special on the Dutch elections of June 2010: blends of possible coalitions: Here!.
News (January 26, 2010)
A new version of FaceBlender is released. It features a new green look and feel, Facebook integration and downloadable external photo libraries. Also a new one-by-one blender tool is included to give more freedom in the images you want to blend.
The Application
FaceBlender is an iPhone/iPod touch application that allows you to blend, not just two, but any number of faces together into a beautiful average. With just a few simple clicks, faces will be automatically scaled and rotated to form a perfect blend.
Use it to:
Import photos from your photo library or take new ones with your iPhone camera, use your contact list or your Facebook friends.
Discover how your average colleague, friend, ex-girlfriend or pet dog looks like.
Blend your face with an average face to see how you look like as being of the opposite sex, or how an asian version of you looks like.
Export the blends you make, so that you can e-mail them to friends or add your creations to your favorite social network profile.
Tag your photos with traits like hair color. Then, use the easy trait selective blender to make blends such as the average female blonde wearing glasses, or smiling colleagues combined with cats.
Extra:
The application comes with free sample databases containing over 100 images so that you can start playing around straight away! (All demo face images originate from stock.xchng and are copyright free with the restrictions given on the website.)
[ Back to top ]
Examples
The average of the male population of a physics group in a Dutch university. Then, the average of several cats that are included in the demo database. Last, averaging the physicists with the cats leads to another 'interesting' result.
The average of several self-portraits of the famous dutch master Rembrandt van Rijn. Then, we also add several self-portraits of Vincent van Gogh (who looks the other way). The result is as puzzling as the young woman/old hag gimmick.
[ Back to top ]
Background
Theories roughly state that people, next to remembering the different faces they see, also keep track of the average of all these faces. Every face that one sees is compared to this average. In general, the more a face resembles the average the more attractive the person will find this face. Of course, this is not a waterproof theory. A blend of 'attractive' faces will give a more attractive average than a blend of 'less attractive' faces. However, the average is usually more attractive than each of the individual faces. In the end though, the average face may be beautiful but it is also relatively boring. That is why people tend to be attracted to more 'exotic' faces. FaceBlender is not intended to help fundamental research that exists in this field, but aims to play around with the ideas. One thing is certain, an average face triggers something in our brains!
[ Back to top ]
Older News (June 3, 2009)
A new version of FaceBlender is released. It features adding photos from your contact list, faster trait selection, name+trait auto-complete feature and smoother use of the manual blender. In addition, several bugs are fixed for iPhone users related to adding photos with your camera; and we fixed several other smaller bugs.
The upcoming release will feature Facebook integration: uploading photos to your Facebook account and adding the profile photos of your friends to your library. The demo library will be made available on a server so that the application itself will be much much smaller. And we have prepared a special one-by-one blender which really makes blending faces a fun pastime.
Older News (April 23, 2009)
A new version of FaceBlender is released which fixes a critical bug for iPhone users and adds localization in French, German, Swedish and Dutch. The upcoming version that is waiting for approval will contain many small updates and will greatly enhance the user-friendliness.
[ Back to top ]
