Meet Julia and Justin.
In this interview, we make acquaintances with Julia and Justin who just recently finished their Master Thesis at CellaVision. Read about why CellaVision was their number one choice when choosing a company.
What was most important to you when it came to the organization for your Master Thesis?
JULIA
The project! I wanted to write my thesis about deep learning, and I thought it would be exciting to do it with a Medtech company to get to know the sector. However, the most important thing for me was to find a research-oriented company that sells a product I believe in, which could offer an interesting ML/AI project. During the spring we were in contact with various companies, but Cellavision was my preferred choice due to what I’d heard from former master thesis students.
JUSTIN
Good question! There were a couple of things that I had in mind when I was looking for a master thesis project. I wanted to work with an organization that I felt makes a meaningful product, meaningful in the sense that it would make a positive impact on society. Ideally, the organization is working with something technically interesting and contemporary (that keeps up with modern research) and cares for the thesis workers, meaning they would not just leave us to solve the issue by ourselves. We were in contact with some other companies in the Malmö/Lund region before choosing CellaVision, there were many companies sounded good, but CellaVision was the one that met the criteria the best and had the most interesting project.
What are you writing your Master Thesis about?
Our Master Thesis is about image reconstruction and compression on cell images. Today, CellaVision uses JPEG to compress cell images, while there may be a loss in quality of the images there is a gain in valuable storage space. A worthy tradeoff. They wanted to investigate a way to reconstruct the JPEG-compressed images, making it possible to keep the gain in storage space while having no visual loss in quality, a.k.a having the cake and eating it too.
By using supervised deep learning models, we have examined and implemented several different models to solve the inverse problem of reconstructing JPEG-compressed cell images.
While JPEG is a great and widely used compression method, it does not take the semantics of the images in consideration, and while there are many different types of blood cells, they are still fairly similar. So, we’re also trying to come up with a way to compress the images via latent space storing, which we think would be a more efficient way of storing cell images.
Have you heard about CellaVision before doing your Master Thesis here?
JULIA
Yes, on several occasions. I knew of other students who did their master thesis projects with CellaVision. Their projects seemed interesting, and all of them recommended us to apply. During the spring I also came in contact with the company during some work fairs at the campus.
JUSTIN
Yes, I have met CellaVision during career fairs at LTH, such as Arkad, Farad, and Teknikfokus, where we have had some great conversations about their products and MedTech in general. Some friends of ours have done their master thesis here and some others that have worked here during the summer. They have all had positive experiences from working here!
Please describe the collaboration between you and your supervisor/mentor at CellaVision.
JULIA
We had two helpful supervisors at CellaVision, Jesper and Martin, who introduced us to the project. From day one they made us feel welcome, and we have had regular meetings with them discussing our progress and the problems we encountered. Considering the global pandemic with a lot of people working from home these days, everything went smoother than expected. We were given the freedom to work independently but were able to ask questions whenever we needed to.
JUSTIN
It’s been great! We have two supervisors who both are highly involved in the project and have come with a lot of great ideas. We speak with them regularly and keep them updated on how it goes.
What is the best thing about doing your Master Thesis with CellaVision?
JULIA
Oh, where should I start? Firstly, we were given an interesting project idea which we were allowed to develop further. Secondly, we’ve had access to a dedicated space for us at the office. Lastly, I have to say it’s the people who work here. They have really made us feel welcome.
JUSTIN
We have a lovely office with a lot of sunlight and great colleagues. There is a gym in the cellar where I’ve been working out, which has saved me on some monthly gym membership costs. It’s a very active office, I’ve worked out with some of the colleagues in the gym.
Could you recommend coming students to do their Master Thesis with CellaVision?
JULIA
Definitely, I am very glad I’ve had this opportunity and I would warmly recommend CellaVision as a great place for writing your thesis.
JUSTIN
Yes absolutely! I agree with Julia!