Deep multi-object tracking, antimicrobial resistance prediction and medical imaging

NAISS 2024/5-250


NAISS Medium Compute

Principal Investigator:

Lennart Svensson


Chalmers tekniska högskola

Start Date:


End Date:


Primary Classification:

10207: Computer Vision and Robotics (Autonomous Systems)

Secondary Classification:

20205: Signal Processing



Our application is intended to provide GPU access to several of our ongoing research projects. These projects focus on a variety of topics where we seek to apply deep learning. We are currently training deep neural networks for 1) antimicrobial resistance prediction, 2) multi-object tracking for self-driving vehicles, and 3) semi-supervised learning with applications in medical image analysis. The application will describe and motivate our needs for GPUs within these three areas. The objective of our project on antimicrobial resistance prediction is to help hospitals select suitable treatments and tests for dealing with resistant bacteria. In the multi-object tracking project, we investigate strategies to improve on the traditional tracking algorithms using deep learning. Finally, our objective in medical image analysis is primarily to analyse CT images for possible atherosclerosis in the coronary arteries, which in turn can predict the risk of myocardial infarction in the future. Annotations are very expensive for medical images, and we are specifically developing self-supervised methods.