Master Thesis at Precise Biometrics
Want to write your Master’s Thesis with us?
Precise Biometrics is constantly searching for dedicated and driven students who want to write their Master thesis in collaboration with us. By doing that you are contributing with valuable knowledge and laying the foundation for further development of our proven biometric technology, which is used millions of times every day.
Are you interested in writing a Master Thesis on one of the following topics, read more about each topic and submit your proposal below:
Project 1
Title: Synthesize Spoof Fingerprint Images
Description: A core part of a biometric fingerprint recognition sensor is spoof detection, i.e. detecting if a presented fingerprint is real or artificial. Creation of spoof detection algorithms requires large amounts of images of real and artificial fingerprints. Collecting data is time consuming and costly. Hence generation of synthetic spoof images would increase the security of the algorithms, as well as decrease the cost and lead time for algorithm creation.
Project 2
Title: Synthesize Fingerprint Images for Different Sensors
Description: When developing fingerprint matching algorithms, a lot of data is needed for both training algorithms and evaluating the performance. Whenever a new sensor is developed, or a modification of an existing sensor is done, new images must be collected. Synthetic fingerprint generation is an interesting new direction for reducing the need of images. However, the generated images must match the characteristics for the sensor at hand, while using a limited number of images from that particular sensor.
Project 3
Title: Detecting Fingerprint Images Outside Training Distribution for Spoof Detection
Description: Synthetic, or spoof, fingerprints can be generated with a large variety of methods, which makes it hard to cover all possible cases. Moreover, it is well known that classification algorithms often behave unexpectedly when fed with data outside the training distribution. Thus, one would like to detect if a presented image is too different from the data that is has been trained on. The goal of this thesis is to explore fast methods for out-of-distribution detection for fingerprint images.
Project 4
Title: Investigate and Develop a Palm Biometric Algorithm
Description: Precise Biometrics has developed state-of-the-art algorithms for fingerprint biometry. An exciting new area is palm prints. Palm prints share many characteristics with fingerprints when it comes to structure and uniqueness. However, to efficiently match palm prints, they must be correctly aligned in three dimensions and stretching/relaxation of the hand must be accounted for. The aim of this thesis is to develop a solution for palm matching using a mix of new and in-house technologies.
Project 5
Title: Event Cameras for Biometric Access Control
Description: An event camera, also known as a dynamic vision sensor, uses independent and asynchronous pixels. They respond to changes of brightness, not absolute level. This makes them extremely fast, with a very large dynamic range, while having a very low bitrate. Precise Biometrics’ product YOUNiQ uses face recognition and liveness validation for allowing access to facilities, for example gyms and offices.
The aim of this thesis is to explore using an event camera for biometric access control.
Project 6
Title: Removal of Moiré Patterns in Fingerprint Images
Description: For under-display sensors a moiré pattern is overlayed on to the fingerprint image. The pattern is created by an alias between the display and the sensor pixels and is very characteristic. It changes with pressure and temperature, and it therefore becomes difficult to filter out. The goal of the Master’s thesis is to remove the moiré pattern with a method that is immune to external changes, and that is fast enough to be implemented in a fingerprint recognition system.