The Fraunhofer Institute for Computer Graphics Research (IGD) is working on new approaches to detect and prevent face-morphing attacks, which present challenges to biometric security systems.
Face morphing involves merging the facial features of two or more individuals into a single image, creating a composite that can deceive both humans and biometric algorithms. The technique is used by criminals to produce counterfeit identification documents, allowing them to cross borders without detection.
To address the issue, Fraunhofer IGD researchers are developing identification algorithms as part of the ATHENE project, Germany’s National Research Center for Applied Cybersecurity. They use deep learning and artificial intelligence technologies, including generative adversarial networks, to simulate potential attack methods. By generating morphed facial images, they aim to train detection systems to recognize and mitigate these types of fraud.
The institute is also focusing on biometric security in embedded systems, such as access control points with limited computing capacity, with their goal being to implement biometric solutions in everyday technologies like smartphones and augmented reality devices. For example, integrating biometric authentication into head-mounted displays used in virtual and augmented reality applications can enhance security by accurately identifying users through the eye region, even with varying eye positions.
Fraunhofer IGD will present its latest research findings at the it-sa Expo & Congress in Nuremberg, slated to run from October 22 to 24. At the event, they will showcase a demo simulating an airport scenario, allowing visitors to experience the challenges of detecting morphed passport photos.
–
September 25, 2024 – by Ali Nassar-Smith
Follow Us