Facial recognition and DNA sequencing technology used to generate facial images from genetic material

Researchers have developed a way to use facial recognition and DNA sequencing technology to create a portrait of an individual from their genetic information, CTV News reports.

A study published last year describes the creation of an algorithm based on comparisons of the sequenced genomes of more than 1,000 people and high-resolution, 3D images of their faces, as well as measurements of eye and skin color, age, height, and weight. Craig Venter, CEO of Human Longevity and co-author of the paper, said the algorithm accurately predicts genetically simple traits, like eye and skin color, with high accuracy, but admits that for many traits, it is limited.

The team was able to generate images for unknown genomes that could be matched to real photos for 8 out of 10 people, but when facial differences were narrowed by restricting the field to a single race, it fell to 5 out of 10.

In theory, the method could be used to aid criminal investigations by enabling investigators to generate a facial image of a suspect based on genetic material left at the scene of a crime. Law enforcement officials in Washington State did just that earlier this year to attempt to generate new leads in a 30 year-old murder case, and police in Calgary engaged a Virginia company specializing in DNA phenotyping to predict the facial appearance of the mother suspected of killing an infant. CTV News reports that the technology has been used “several” other times in Canadian homicide investigations.

Critics of the method say that it appears to generate images based on averages of gender and ancestry, rather than specific characteristics, according to CTV News.

Article Topics

biometrics  |  dna  |  facial recognition  |  forensics  |  law enforcement