Researchers develop new technique for matching DNA to faces
A team of Belgian and American researchers have developed an algorithm that matches DNA to potentially-matching facial images, which they suggest may have possible biometric forensic applications.
A team led by engineers at KU Leuven published their findings in Nature Communications, including that checking if faces could match DNA yields better results than “going from DNA to face,” as KU Leuven Electrotechnical Engineer Peter Claes puts it.
Facial appearance is determined in part by thousands of genes, but also by living conditions, including the food someone eats. The researchers say this makes it unlikely that DNA alone could ever be used to accurately predict an individual’s face. The technology is expected to become more accurate, however, as more genes are identified. It also could have forensic value without being able to make a positive identification.
“This method mostly helps rule people out. In practice, we don’t usually get any further than a sort of reference face, such as ‘a European male,’” says Claes. “That’s not much use to a forensic investigator. Unfortunately, while we are learning about more and more genes that determine certain aspects of our face, this does not yet sufficiently translate to a better match between the predicted face and the faces in the database.”
In addition to matching, the technology could be used to sketch an approximation, which the researchers say could then be matched against a database. U.S. police in Oregon and New York have allegedly used sketches with facial recognition systems.
MIT researchers recently unveiled a neural-network based model for approximating facial appearance from vocal characteristics.