June 30, 2014 -
Researchers at the University of Central Florida (UCF) have developed a facial recognition tool that they say can be used to quickly match photos of children with their biological parents, along with determining photos of missing children as they get older, as recently presented at the IEEE Computer Vision and Pattern Recognition conference in Columbus, Ohio.
Led by graduate student Afshin Dehghan, the UCF’s Center for Research in Computer Vision team initially launched the project using a database that holds more than 10,000 photos of celebrities, politicians and their children.
Dehghan told UCFToday his team is trying to determine whether the facial recognition tool could assess this data to objectively answer questions like “‘Do children resemble their parents?’ ‘Do children resemble one parent more than another?’ and ‘What parts of the face are more genetically inspired?'”.
The tool could help law enforcement and families find children who have been missing for a long time, said Ross Wolf, the associate dean in the college of health and public affairs and an associate professor of criminal justice at UCF.
Wolf adds that law enforcement is already using face recognition technology, however, the new technology developed by Dehghan`s team could potentially improve on its limitations by identifying the same facial features in photos over time.
As part of their research, Dehghan and his team are also looking at how factors like age and ethnicity can ultimately impact the resemblance of facial features.
Meanwhile, new facial recognition software developed by medical researchers and Professor Andrew Zisserman from Oxford University’s Department of Engineering Science could help doctors diagnose rare genetic disorders, according to a separate report by Independent.co.uk.
The new software, which is similar to technology used in new handheld cameras, is based on the analysis of thousands of images of previously diagnosed patients.
As a result, the software effectively “learns” what the essential and non-essential facial features are in terms of making a diagnosis.
The tool will also be able to match images of patients with unknown disorders who possess similar facial features and skull structures in a separate group, which could in turn help doctors to identify new disorders, as well as the DNA variations that cause these disorders.
By using advanced computer vision technology, the software will “learn” from a deposit of 3,000 patient photographs taken from public and clinical databases.
Dr Christoffer Nellaker, of the Medical Research Foundation’s Functional Genomics Unit at Oxford, said the technology`s ability to diagnose a rare genetic disorder is a crucial advancement for doctors and patients.
He added that in the future a doctor would be able to take a photo of a patient using a smartphone and run the computer analysis to rapidly determine what kind of genetic disorder he or she may have.
As mentioned in a recent guest post on BiometricUpdate.com, facial recognition technology has seen a recent backlash in the news with numerous wormhole-like theories emerging that the technology leads to mass surveillance, the destruction of anonymity, and changing the way people behave in public.