In December 2024, Robert Nichols successfully defended his MSc thesis with the title „A Composite Framework for Context-Based Data in Differential Morphing Attack Detection“.
Abstract:
Morphing attacks in the border control scenario remain a threat to public security. This makes them an area of continued significant scientific interest. This work addresses one of the main limiting factors in biometric machine- and deep-learning research, specifically morphing attack detection: the lack of available, high-quality, and accurately labeled data spanning up to 10 years of sample age intervals. To this end, a novel data collection approach is proposed that integrates multiple state-of-the-art computer vision and natural language processing models with technologies that constitute the World Wide Web, with the primary data source being live and archived news programs from German public television, and accurate labeling supplied by large-scale public knowledge graphs. State-of-the-art image morphing methods are applied to create a contemporary database for developing novel morphing attack detection systems. Additionally, using this new database, a preliminary novel approach to morphing attack detection is explored that follows recent findings from the object detection task, which incorporates additional information to boost detection performance. The results indicate a potentially fruitful interdisciplinary research area and promote a multi-modal view of traditionally unimodal vision tasks.