Meet our new Master’s thesis students Anna, Tove and Lukas!
How did you become aware of Precise Biometrics, and what attracted you to do your project with us?
Anna and Tove: We stumbled upon Precise Biometrics during our search for a suitable company to collaborate with for our master’s thesis. Initially, we were unfamiliar with the company, but the more we read about the company the more interested we got since we both have a large interest in machine learning. In September, we had a meeting with Anders Olsson where we explained our interests in AI and business development. The meeting ended with a proposition for a very interesting project that both aligned perfectly with our and Precise’s interests. This meeting was essentially what attracted us to do our project with Precise. We are thrilled to be here and to see what the project will result in.
Lukas: I am very grateful to my friend Julia who compiled a large list of potential companies for a Master’s Thesis in Lund regarding AI. Amongst the numerous candidates Precise Biometrics stood out to me as a leading company where my work had the potential to have an actual impact.
What are the main research questions you aim to answer through your thesis?
Anna and Tove: Our objective is to identify promising areas within healthcare or medtech where Precise could thrive and extend their business presence. Additionally, we’ll explore the opportunities and challenges present in the biomedical market and find strategies for Precise to capitalize on opportunities while mitigating potential threats.
Lukas: I will be looking at developing a model for classifying palm prints. There are many steps in the process to consider: datasets, preprocessing, architecture, and analysis. Furthermore, there are additional considerations like security, privacy, and multi-modal inputs that I hope to widen the scope of.
How do you believe your thesis contributes to the existing knowledge in AI and biometric technology?
Lukas: Currently, there is limited research in using large datasets of hands for palmprint classification. Many articles have only a singular dataset which they test their results on. Although the results they show are good, I hope to incorporate many different datasets and provide a model which can work well on datasets containing a vast collection of people.
In what ways do you believe your research could be expanded or improved upon in the future?
Anna and Tove: We’re convinced that AI holds great potential in shaping the future of healthcare and medtech, with numerous areas ready to benefit from its integration. Our project aims to explore these potential applications of AI in healthcare and lay the groundwork for its broader utilization in the field.
Lukas: The current data I am using does not have any good variation in lighting or poses which could lead to performance issues in real world applications. The theory might look good but when subpar cameras are involved the model might fail.
What are the practical implications of your research findings, and how do you envision them being applied in real-world settings?
Lukas: Currently biometrics are a core part of phone usage and functionality during identification. They are considered secure enough to be used in payments. I see no reason that palmprints cannot reach the same level, given sufficient work and time. In the future a person might be able to enter their office just by holding up their palm or collect their package from the post office by scanning their hand.