Researchers at the University of Washington have developed a predictive age progression app that can accurately predict what a child will look like as an adult, up to the age of 80. Aside from offering up some predictive fun, the software could play a crucial role in solving missing persons cases. The software algorithm, developed by a team led by professor Ira Kemelmacher-Schlizerman, is claimed to be able to do a better job than a human.
The software uses a database of thousands of faces grouped by age and gender. It determines the average pixel distribution of different facial features and calculates how those features change as a person ages.
It starts by taking a baby picture as input. It then corrects for tilts, turned heads and lighting before applying age-related changes to the baby photo to predict how he or she will look will as an adult.