May 4, 2015 – Law enforcement and national security agencies could benefit from an Oak Ridge National Laboratory technology able to determine a person’s age, race and gender with high fidelity. “Normally, computers estimate age by looking for wrinkles or estimate gender by looking at specific two-dimensional distances or 2-D texture,” said Ryan Tokola of ORNL’s Imaging, Signals and Machine Learning Group. ORNL’s system allows users to employ the same set of features to estimate age with an error of less than five years, gender with 89 percent accuracy and race with 99 percent accuracy. This is the first work to accomplish this based solely on the 3-D geometry of a face. Tokola will present this work at the International Conference on Biometrics in Thailand May 19-22.