Advanced Seminar / Masterseminar WS 2018/19
The Advanced Seminar will focus on different topics related to Biometrics
Seminar will start at 10.10.2018 at 16:00 in room D19/2.03
The seminar will have only few fixed class meetings. Besides these meetings, additional appointments shall be arranged individually and on demand. The schedule of this seminar can be found in the online booking system (OBS).
Each group’s term paper has to be prepared using Springer’s LNCS template (Word or LaTeX) with a length of 8 – 10 pages, not counting references and clear marked appendices. The final presentation will be 45 minutes per group + 15 minutes discussion of the results. A grade will be given based on the term paper and final presentation as defined in the module description.
Student groups (1-2 persons) may select their own topic or choose a topic from the list below. Detailed topic descriptions and starting resources can be found in a separate document.
- Synthetic Finger-Vein Data Generation and Evaluation: Synthetic generation of biometric data enables evaluations on scale normally impossible for manually collected research datasets. In this project, a finger-vein data generator will be used to generate a large biometric database, and conduct an evaluation in accordance with the ISO/IEC standard methodology.
- Vein Image Quality Assessment and Enhancement: Vein recognition systems gain popularity in real life applications, for example many Japanese banks use this biometric modality to secure the ATM access. The quality of captured samples is a critical aspect in biometric systems. In this project, a survey will be conducted, with emphasis on the impact of image quality on vein recognition systems, metrics for assessment of vein image quality, as well as existing approaches for quality enhancement of vein images.
- Online Signature Acquisition: Typical online signature acquisition protocol involves the data subjects having to repeatedly type their signature a number of times. In this project, it will be investigated, whether there are differences between successive acquisitions, for instance due to the data subjects becoming more consistent, or bored, quick, and careless in their typing.
- Application Scenarios of Machine Learning within the Field of Touchless Fingerprint Recognition: Machine learning have reached a many of application scenarios in recent years. Several approaches have been proposed at different stages of touchless fingerprint recognition. In this work the touchless fingerprint processing stages should be extracted, existing neuronal approaches discussed and new possibilities suggested.
- Usability Issues of Slap Hand Acquisition Schemes using Smartphones: Especially slap hand acquisition offers a higher performance compared to single finger approaches because of the availability of more biometric features. Up to now there are a few usability studies on this topic. The task is to define different framework conditions for the acquisition and evaluate its usability with respect to image quality.
- Finger Movement Estimation: One of the advantages of mobile touchless fingerprint acquisition is the ability to easily acquire four fingerprints in one impression. During an acquisition the slap hand is presented to the camera which contains a certain movement. The task is to detect the movement of the slap hand and single fingers and estimate their direction and speed.
- Issues and Challenges of Touchless Fingerprint Database Collection: Mobile touchless fingerprint recognition can be implemented on a commodity devices like smartphones or webcams. For this reason different processing workflows and test samples are needed. During the development of fingerprint recognition workflow a database of biometric samples is required to evaluate the implemented algorithms. With a given acquisition scenario the task is to specify framework conditions and metrics in order to establish a database of high quality samples and show its usability.