Student work and practical projects

Biometrics is still a relatively young technology with great potential for development, which gives rise to a wide range of fundamental questions that still need to be investigated. In close co-operation with the BSI, the ISF offers support for final theses and practical projects.

Current topics for theses and practical projects:

Counterfeit detection technologies (Presentation Attack Detection, PAD):

• Non-contact measurement of the pulse using a camera
• Material characterisation with a 3D camera using scattering properties
• 3D counterfeit detection
• Skin detection near-infrared PAD
• Make-up detection
• Deepfakes (Face, Voice)
• Analysing biometric data
  • Biometric background systems
• Synthetic characters
• Test equipment production
  • Material analysis
  • Production of counterfeits using 3D printing

You can also find more information on the topic of theses in the field of biometric procedures on the BSI website. In addition to the topics listed, we also offer the opportunity to submit your own topic suggestions.

If you are interested in completing your Bachelor's thesis, Master's thesis or practical semester at the BEZ, you can contact the project managers directly (contact BEZ project management).

Completed work and projects at BEZ:

• Presentation Attack Detection (PAD):
  • Presentation-Attack-Detection based on the pupil light reflex (project study)
  • Materials science fundamentals of skin and biometric forgeries for facial and fingerprint biometrics
  • Light scattering as a material property for differentiating between human skin and biometric forgeries
  • Conception and validation of the integration of counterfeit detection based on NIR spectroscopy in a fingerprint scanner
  • Conception, realisation and validation of the integration of counterfeit detection based on NIR spectroscopy in an ultrasonic fingerprint scanner
  • Automated monitoring for biometric applications using 3D position analysis of humans
  • Development of a test setup and realisation of attacks on a radar sensor used in the automotive sector
• Deepfake:
  • Toolbox development for the implementation and fusion of deepfake detection methods to evaluate generalisability
  • Audio deepfake detection: Relevant factors and influences of the data on generalisability
• Template protection procedure:
  • Template protection procedures in the context of finger vein biometry
  • Investigation of template protection methods for SWIR facial biometry
  • Evaluation of the security of hardened fuzzy vault methods
  • Analysis of a fuzzy vault scheme extended by minutiae angles
• Face recognition:
  • Development and evaluation of test criteria for biometric face recognition in eGates
  • Analysis and evaluation of spoofing attacks in automatic face recognition
  • Localisation and verification of human faces in near-infrared images
• Biometric background systems:
  • Functional and security-related expansion of a data protection-compliant analysis framework for evaluating biometric systems in the Biometric Evaluation Centre (BEZ)
• 3D cameras:
  • Metrics and characterisation of 3D time-of-flight cameras for facial biometric applications
• OCT fingerprint scanner (Optical Coherence Tomography):
  • Design and realisation of a sensor system for monitoring the unattended operation of an OCT fingerprint scanner in kiosk mode
  • Extraction of biometric features from FD-OCT finger scans
  • Investigation of a layer recognition system for extracting the inner and outer fingerprint from OCT volume data
• Vein biometrics:
  • Design of a processing chain for vein extraction
  • Vascular feature extraction methods in biometrics - development, analysis and testing
  • Development of an evaluation and development framework for finger vein detection methods based on the Fingervein V1.5 acquisition software
  • Improving the image quality of vein scanners for vascular biometry
• Morphing:
  • Morphing detection with low-cost 2D and 3D cameras compared in practice
• Other:
  • Design and development of a framework for analysing various biometric sensors from different manufacturers
  • Modelling and virtual representation of artificial replacement teeth in patient photos for a before-and-after comparison
  • Identification of a vehicle driver based on dynamic driving behaviour data
  • Semantic image segmentation by combining visible and near-infrared channels with depth information