Examensarbeten för masterexamen


Senast publicerade

Visar 1 - 5 av 1991
  • Post
    Robust Medical Image Analysis using Privileged Information
    (2024) Chakrabarti, Apala; Chalmers tekniska högskola / Institutionen för elektroteknik; Häggström, Ida; Johansson, Freddrik
    Abstract Domain adaptation is a crucial task in medical diagnosis and treatment planning, as it enables models trained on large, labeled datasets to be effectively applied to smaller, domain-specific datasets. This is particularly challenging due to data scarcity and shifts in data distribution. Privileged Information (PI), such as binary attributes or bounding boxes, has the potential to improve machine learning models’ adaptability across diverse domains. This study aims to investigate the role of PI in domain adaptation for medical image classification. The results of this experiment indicate that integrating PI led to increased accuracy and stabilized prediction accuracies. Furthermore, the findings affirm the importance of both the quantity and correlation of the PI provided and its correlation with output labels in enhancing model performance, thereby supporting the fundamental principles of domain adaptation. Moreover, the study underscores the significance of strategically considering PI attributes during model training to achieve stable output accuracy and effectively mitigate domain shift. This comprehensive study will help improve diagnostic accuracy in various domains, especially healthcare, which can lead to more effective treatments and better patient outcomes.
  • Post
    Tracking players and ball in football videos
    (2024) Ganelius, Hugo; Humayun, Jhanzaib; Chalmers tekniska högskola / Institutionen för elektroteknik; Svensson, Lennart; Lennart
  • Post
    Football Analysis in VR - Texture Estimation with Differentiable Rendering and Diffusion Models
    (2024) Anjou, Filip; Ekström, Albin; Chalmers tekniska högskola / Institutionen för elektroteknik; Svensson, Lennart
  • Post
    Human centric virtual driven product development - Using FE simulation to compare passive versus active muscle activity for human body model development
    Tjernberg, Sigrid; Chalmers tekniska högskola / Institutionen för elektroteknik; Fhager, Andreas; Mendoza Vazquez, Manuel
    Abstract The aim of this thesis was to gather information about active and passive muscle tissue around the pelvis and use it to simulate the effect on a human body model (HBM). A literature review was conducted in order to find muscle tissue parameters that could be used in the simulation of muscles around the pelvis. Three studies were chosen to compare muscle tissue parameters, the last of which included muscle tissue parameters for both passive and active muscle tissue. The literature review was also conducted to investigate the effects of body mass index (BMI) on the muscle tissue properties. Then two models were designed in order to simulate in vivo and in vitro muscle tissue tests which were used for the first sets of simulations. The first simulations were designed to be done quickly so that several different muscle tissue parameters could be tested effectively and compared. Part of this comparison was to check the material model stability in the range of relevant strains. The last step was to implement the muscle tissue parameters in a HBM and see how it affected the model. The results showed that there is a lack of information within the field and a lot of misinformation where bad values were being used, however a few studies which were deemed as good which had parameters which could be used in ABAQUS FEA were summarized in this thesis. All these muscle tissue parameters were tested on the cylinder designs which resulted in force-displacement plots which could be compared to each other. Active and passive muscle material parameters were then applied to a HBM which resulted in two different models which were compared. The conclusion of the thesis showed that active versus passive tissue does affect the HBM.
  • Post
    Distributed DC/DC converter system - Maintaining voltage stability and efficiency optimization while regulating several converters
    (2024) Johansson, Markus; Tepp, Isabella; Chalmers tekniska högskola / Institutionen för elektroteknik; Thiringer, Torbjörn; Larsson, Torbjörn; Singh, Aditya Pratap
    Abstract This thesis is about the investigation of a new concept regarding the supply of the low voltage side in battery electrical vehicles. The concept is about supplying the low voltage side with a multilevel converter (MLC). A control strategy was tested that is based on distributing different set voltages to each converter. The system was meant to regulate itself with no other information other than the input voltage and output voltage. The system needs to fulfil the requirements of LV124. The test sequences performed showed that a system of three converters can handle power loads ranging from approximately 20−600 W, while still fulfilling the LV124 requirement. The test sequences were performed on a system of converters with different set voltages as well as with the same set voltage. The result was that the test case with converters on the same set voltage performed better regarding both the overall efficiency as well as the voltage stability. However, the theory behind this thesis and the simulated models counter this result, meaning that a system with different set voltages would improve the overall efficiency. This unpredicted result can be concluded to be the cause of a hardware that is still in the early stages. With further development of the hardware to improve efficiency and control strategy to improve voltage stability of the system, this MLC system could possibly supply the low voltage side.