Edge-Based Self-supervision for Semi-supervised Few-Shot Microscopy Image Cell Segmentation
Date
2022
Authors
Dawoud, Y.
Ernst, K.
Carneiro, G.
Belagiannis, V.
Editors
Huo, Y.
Millis, B.A.
Zhou, Y.
Wang, X.
Harrison, A.P.
Xu, Z.
Millis, B.A.
Zhou, Y.
Wang, X.
Harrison, A.P.
Xu, Z.
Advisors
Journal Title
Journal ISSN
Volume Title
Type:
Conference paper
Citation
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2022 / Huo, Y., Millis, B.A., Zhou, Y., Wang, X., Harrison, A.P., Xu, Z. (ed./s), vol.13578, pp.22-31
Statement of Responsibility
Youssef Dawoud, Katharina Ernst, Gustavo Carneiro, Vasileios Belagiannis
Conference Name
International Workshop on Medical Optical Imaging and Virtual Microscopy Image Analysis (MOVI) (18 Sep 2022 - 18 Sep 2022 : Singapore)
Abstract
Deep neural networks currently deliver promising results for microscopy image cell segmentation, but they require large-scale labelled databases, which is a costly and time-consuming process. In this work, we relax the labelling requirement by combining self-supervised with semi-supervised learning. We propose the prediction of edge-based maps for self-supervising the training of the unlabelled images, which is combined with the supervised training of a small number of labelled images for learning the segmentation task. In our experiments, we evaluate on a few-shot microscopy image cell segmentation benchmark and show that only a small number of annotated images, e.g. 10% of the original training set, is enough for our approach to reach similar performance as with the fully annotated databases on 1- to 10-shots. Our code and trained models is made publicly available https://github.com/Yussef93/EdgeSSFewShotMicroscopy.
School/Discipline
Dissertation Note
Provenance
Description
Access Status
Rights
© 2022 The Author(s), under exclusive license to Springer Nature Switzerland AG