Project 1 Exploring Quantum Machine Learning through Kernel Methods

Team: Carl Dahlin, Axel Evelius
Supervisors: Vahid Azimi Mouslou, Prashant Singh
Contact: calle.dahlin@live.se, aom.noheve@gmail.com,
vahid.azimi-mousolou@physics.uu.se, prashant.singh@scilifelab.uu.se
Poster
Project report

Project 2 A Modified Pipeline for Face Recognition

Face recognition consists of four distinct stages. First a face is detected. Secondly, anchor points are found so that the face can be aligned, in order to have the exact same position for all faces. Next a feature vector is computed from the aligned face. In the final stage, verification is performed, usually by computing the cosine similarity between feature vectors.

Pipelines that allow to use different state of the art algorithms for these three steps exist, such as DeepFace, which allows to use a variety of software for all stages in the pipeline. InsightFace, handles profile faces much better, but has the disadvantage of require a parameter setting for the size of faces in images. But on the other hand it is fast. FaceTorch also handles profile faces pretty good, but is generally slower. However, it has the advantage of finding the optimal parameter setting by itself.

The purpose of this project is to evaluate different methods for all stages of the pipeline and create a pipeline for research purposes, where algorithms can easily be inserted. For an example, several pipelines use ArcFace to create peace feature vectors, but newer ones exist, such as MagFace, CurricularFace, ElasticFace and AdaFace. These should be evaluated with respect to intra- and inter-similarities between persons of different age groups, using recent datasets. Another important thing would be to use the face detection of FaceTorch or DeepFace and then run InsightFace, so that no hard coded parameters are necessary.

One important task is to make the computations efficient when working with large datasets and some parallel implementation is therefore necessary.

Team: Tyra Axén, Hanna Ericson
Supervisor: Anders Hast
Contact: tyra_axen@hotmail.se, Hanna.ericson.9834@student.uu.se,
anders.hast@it.uu.se
Poster
Project report

Project 8 Finite Element Approximation of the Kuramoto–Sivashinsky Equation

The goal of the project is to find approximate solutions to the Kuramoto–Sivashinsky equation, which is a partial differential equation involving Laplace, nonlinear, and biharmonic (fourth-order derivative) terms. This equation is used to model and study complex spatiotemporal dynamics and pattern formation. Its main applications are in systems where instabilities lead to chaotic or turbulent behavior, often occurring in areas related to fluid dynamics, combustion, and plasma physics. The equation will be solved using the FEniCS project, which is an open-source finite element library.

Project 9 Shock formation study using the compressible Euler equations

This project focuses on the numerical simulation of shock formation through the compressible Euler equations. We implement a finite element approximation using the FEniCS framework, incorporating a nonlinear artificial viscosity technique to stabilize the numerical scheme. The solver is applied to various challenging gas dynamics problems, including the formation of shocks and discontinuities. These simulations allow for a deeper exploration of complex flow behaviors and the mechanisms that lead to shock development in compressible fluids.

Team: Majd Chahin, Robin Dymér, Anna Johansson, Alexander Sundlöf Edman
Supervisor: Murtazo Nazarov
Contact: mjedshahin2000@gmail.com, robindymer1@gmail.com, annjoh004@gmail.com, alexander.sundlof@gmail.com,
murtazo.nazarov@uu.se
Poster
Project report

Project 10 Multi-objective Linear Programming for Radiotherapy

Team: Emil Forsberg, Carl Agélii, Viktor Lindström, Clas Zachrisson
Supervisors: Daniel Hernández Escobar, Jens Sjölund
Contact: Emil_Forsberg@outlook.com, calle.agelii01@gmail.com, Viktorrlindstromm@gmail.com, clas.zachrisson.6871@student.uu.se,
daniel.hernandez.escobar@it.uu.se, jens.sjolund@it.uu.se
Poster
Project report

Project 13 Deep Learning for Image Analysis of 3D Microtumor Cultures: Applying CNNs and Data Augmentation to Study Anti-Tumor Treatment Responses in Microwell Platforms

Team: Hugo Toll, John Forslund, Josefine Hörnell, Viktor Bergstedt
Supervisors: Patrick Sandoz, Carolina Wählby
Contact: hugo.toll@live.se, forslundjohn@hotmail.com, josefine.hornell@gmail.com, viktor.bergstedt@it.uu.se,
psandoz@kth.se, carolina.wahlby@it.uu.se
Poster
Project report

Project 16 Is quantity a quality of its own? -- Optimising test milkings for precision dairy farming

Project 17a Efficient deterministic pricing of basket options on many underlying assets

Project 17b Efficient deterministic pricing of basket options on many underlying assets

Team: David Erik Hammar, Erik Björklund, Noah Waltman
Supervisor: Elisabeth Larsson
Contact: david.erik.hammar@gmail.com, erik.blund@gmail.com, noahwaltman123@gmail.com,
elisabeth.larsson@it.uu.se
Poster
Project report

Project 18 Uncertainty Quantification from Frozen Embeddings

Team: Stina Fredriksson, Max Andersson, Edvard Glockner
Supervisor: Prashant Singh
Contact: stina.ida.fredriksson@gmail.com, maxandersson8@outlook.com, edward.glockner5@gmail.com,
prashant.singh@scilifelab.uu.se
Poster
Project report

Project 21 Analysing data from the Internet Microscope: Classification of home network traces

Team: Niclas Svensson, Patrik Grimlund
Supervisor: Mike Hazas
Contact: niclasdan@gmail.com, patrik.grimlund@gmail.com,
mike.hazas@it.uu.se
Poster
Project report

Project 22 Deep Learning Based Focus Interpolation for Microscopy

Team: Olle Edgren Schüllerqvist, Olof Jardfelt, Jonathan Arvin, Axel Sigstam
Supervisor: Joakim Lindblad
Contact: olle.edgrenschullerqvist.6429@student.uu.se, olof.jardfelt@gmail.com, Jonathanarvin@outlook.com, axel.sigstam@gmail.com,
joakim@cb.uu.se
Poster
Project report

Project 30 Inventory Optimisation in Vending Machines

Team: Linus Westerberg
Supervisor: Aria Assadi
Contact: linuswesterberg@hotmail.com,
aria@impactsolution.se
Poster
Project report

Project 33 Using Image Analysis to Determine the Product Map of Vending Machines

Team: Sanni Lounela, Johan Stenborg
Supervisor: Aria Assadi
Contact: sanni.lounela@gmail.com, Johan.Stenborg995@gmail.com,
aria@impactsolution.se
Poster
Project report

Project 34 Unsupervised learning for untargeted spatial analysis of mass spectrometry imaging

Team: Gustav Malmsten, Alicia Karlsson Spongh, Zezheng Zhang, Marton Bidlek
Supervisors: Orcun Göksel, Ingela Lanekoff
Contact: gustavmalmsten@gmail.com, alicia.spongh@gmail.com, zezheng.zhang.8050@student.uu.se, bidlekmarton@gmail.com,
orcun.goksel@it.uu.se, ingela.lanekoff@kemi.uu.se
Poster
Project report

Project 35 Mixed-Integer Programming Model for Optimal Hardware Activation Code Recommendation

Team: Olle Virding, Love Leoson, Alexander Eketrapp
Supervisor: Erik Cervin Edin
Contact: olle.virding@gmail.com, love@leoson.nu, alexander.eketrapp@gmail.com,
erik.cervin.edin@ericsson.com
Poster
Project report