New software that can mature babies' faces to predict what they will look like when they are 80 has been developed by researchers at the University of Washington.
Because our appearance changes dramatically as we grow into adulthood, predicting what a baby will look like in old age was considered a challenge by software developers.
When the researchers set up an experiment using baby photos of 82 real people to test the software, they found that people matched digitally rendered images of the adults with the correct baby image as often as they connected them with the real-life photos.
In fact, they are "so convincing that people can't distinguish them from reality", said Steven Seitz, a professor of computer science.
The software relies on the images of children and adults from the internet, and makes use of added information such as the age and gender of the subject. Using an algorithm, the programme matches age and gender information, identifying the average pixel composition for matching photographs and predicting how each face will change.
Even more useful perhaps is the time in which this software can accomplish its task. Used on an ordinary computer, it generates images in about 30 seconds.