Xu, Mou-Cheng; Zhou, Yukun; Jin, Chen; Groot, Marius de; Alexander, Daniel C; Oxtoby, Neil P; Hu, Yipeng; Xu, Mou-Cheng; Zhou, Yukun; Jin, Chen; Groot, Marius de; Alexander, Daniel C; Oxtoby, Neil P; Hu, Yipeng; Jacob, Joseph; - view fewer (2022) Bayesian Pseudo Labels: Expectation Maximization for Robust and Efficient Semi-Supervised Segmentation. In: Proceedings of 25th International Conference on Medical Image Computing and Computer Assisted Intervention 2022. Springer Nature: Cham, Switzerland.

Preview

Text 2208.04435v1.pdf - Accepted Version Download (1MB) | Preview

Abstract

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 .

Download activity - last month

Export as JSONXMLCSV

Download activity - last 12 months

Export as CSVXMLJSON

Downloads by country - last 12 months

Export as JSONCSVXML

Archive Staff Only

View Item