Joris Roels

Joris Roels obtained his Masters degree in Mathematical Computer Science in 2014 at Ghent University. In 2015, he obtained an FWO-SB grant and started working on his PhD at Ghent University, in collaboration with the VIB institute. The project was focused on image restoration and segmentation in 3D electron microscopy (EM) data based on statistical frameworks and machine learning algorithms. During his PhD, he worked as a visiting scientist in the IAL research group of prof. Fred Hamprecht (University of Heidelberg) and the research group of dr. Anna Kreshuk (EMBL) on domain adaptive segmentation algorithms for EM. He also co-developed DenoisEM: a GPU accelerated ImageJ plugin that allows for interactive and scalable image denoising. He obtained his PhD degree in 2019 and has since then been working as a postdoctoral researcher in the DAMBI research group, headed by prof. Yvan Saeys and the VIB Bioimaging Core, headed by dr. Saskia Lippens. His current work aims to reduce time-consuming and expensive labeling efforts in biological image analysis. This involves developments in domain adaptation, weak supervision, active learning and semi-supervised learning.