Von Chamier, Lucas; (2023) Creating a platform for the democratisation of Deep Learning in microscopy. Doctoral thesis (Ph.D), UCL (University College London).

Preview

Text von Chamier_10163468_Thesis_sigs_removed.pdf Download (14MB) | Preview

Abstract

One of the major technological success stories of the last decade has been the advent of deep learning (DL), which has touched almost every aspect of modern life after a breakthrough performance in an image detection challenge in 2012. The bioimaging community quickly recognised the prospect of the automated ability to make sense of image data with near-human performance as potentially ground-breaking. In the decade since, hundreds of publications have used this technology to tackle many problems related to image analysis, such as labelling or counting cells, identifying cells or organelles of interest in large image datasets, or removing noise or improving the resolution of images. However, the adoption of DL tools in large parts of the bioimaging community has been slow, and many tools have remained in the hands of developers. In this project, I have identified key barriers which have prevented many bioimage analysts and microscopists from accessing existing DL technology in their field and have, in collaboration with colleagues, developed the ZeroCostDL4Mic platform, which aims to address these barriers. This project is inspired by the observation that the most significant impact technology can have in science is when it becomes ubiquitous, that is, when its use becomes essential to address the community’s questions. This work represents one of the first attempts to make DL tools accessible in a transparent, code-free, and affordable manner for bioimage analysis to unlock the full potential of DL via its democratisation for the bioimaging community.

Download activity - last month

Export as CSVXMLJSON

Download activity - last 12 months

Export as CSVXMLJSON

Downloads by country - last 12 months

Export as JSONXMLCSV

Archive Staff Only

View Item