Use of a Neural Network to Identify Man-made Structure in Millimeter-Wave Images for Security Screening Applications

Events in the past few years have heightened security concerns necessitating the development of more advanced methods for detecting potential threats being carried on individuals. One approach is to use imaging methods that see through clothing to find potentially threatening objects being concealed by individuals on their person. This sparks obvious privacy concerns. This paper describes one technique based on neural networks and Fourier features applied to active millimeter-wave imagery that finds man-made structure potentially indicating a threat without compromising personal privacy.