Science News – November 21, 2013

Although the widely held view is that computer recognition of people depends almost entirely on faces, a new study by researchers at the National Institute of Standards and Technology (NIST) and the University of Texas at Dallas reveals that more information, such as body shape,

could enhance identity recognition

.

 

Jonathon Phillips, who works on biometrics at NIST, explains that “[f]or twenty years, the assumption in the automatic face recognition community has been that all important identity information is in the face. These results should point us toward exploring new ways to improve automatic recognition systems by incorporating information about the body beyond the face.”

 

In this series of experiments, researchers showed study participants pairs of images of people; the subjects’ faces and bodies were included in the images. Researchers tracked the eye movement of study participants and found that that there is a reliance on viewing a person’s body to aid the identification process. This suggests that automatic face-recognition computer programs could also benefit from including the body of individuals during the recognition process.

 

This research demonstrates that our assumptions — even those embedded in forensic disciplines — need to be verified by scientific research to understand their reliability. In this case, expanding what information is included in computer recognition programs could lead to a better, more accurate technique.