Biometrics S-MAD (ING9101)

An elective undergraduate course focused on biometric recognition and presentation attack detection, with particular emphasis on face and iris morphing attacks. The course combines theoretical foundations and hands-on implementation of state-of-the-art and classical biometric techniques, covering both deep learning approaches and handcrafted feature descriptors for morphing detection and Single-image Morphing Attack Detection (S-MAD). Students develop practical skills in designing, training, and evaluating biometric systems using real-world datasets.

Technologies and Tools

• Programming language: Python
• Deep learning frameworks: PyTorch
• Computer vision libraries: OpenCV, NumPy, Matplotlib, scikit-image
• Machine learning: scikit-learn
• Biometric topics: Face recognition, iris recognition, morphing attacks, Single-image Morphing Attack Detection (S-MAD), biometric security, performance evaluation
• Methods and techniques: Convolutional Neural Networks (CNNs), transfer learning, handcrafted feature descriptors (e.g., LBP, HOG, SIFT), feature extraction, image preprocessing, classification, and biometric evaluation metrics