This paper concerns pseudo labelling in segmentation. Our contribution is fourfold. Firstly, we present a new formulation of pseudo-labelling as an Expectation-Maximization (EM) algorithm for clear statistical interpretation. Secondly, we propose a semi-supervised medical image segmentation method purely based on the original pseudo labelling, namely SegPL. We demonstrate SegPL is a competitive approach against state-of-the-art consistency regularisation based methods on semi-supervised segmentation on a 2D multi-class MRI brain tumour segmentation task and a 3D binary CT lung vessel segmentation task. The simplicity of SegPL allows less computational cost comparing to prior methods. Thirdly, we demonstrate that the effectiveness of SegPL may originate from its robustness against out-of-distribution noises and adversarial attacks. Lastly, under the EM framework, we introduce a probabilistic generalisation of SegPL via variational inference, which learns a dynamic threshold for pseudo labelling during the training. We show that SegPL with variational inference can perform uncertainty estimation on par with the gold-standard method Deep Ensemble.
@article{arxiv.2208.04435,
title = {Bayesian Pseudo Labels: Expectation Maximization for Robust and Efficient Semi-Supervised Segmentation},
author = {Mou-Cheng Xu and Yukun Zhou and Chen Jin and Marius de Groot and Daniel C. Alexander and Neil P. Oxtoby and Yipeng Hu and Joseph Jacob},
journal= {arXiv preprint arXiv:2208.04435},
year = {2022}
}