Deep co-training for semi-supervised image segmentation
作者：
Highlights：
• Semi-supervised semantic segmentation based on an ensemble of deep learning models.
• All models are trained jointly in a co-training setting.
• Coherence among models is enforced by minimizing the Jensen-Shannon divergence of the probabilities distributions on the unlabeled samples.
• Diversity among models is preserved by enforcing similarity between the prediction of a model on an unlabeled sample and the adversarial prediction of another model on the same sample.
• Results on ACDC dataset and SCGM dataset show the capability of our model to outperform previous ensemble approaches by a significant margin.
摘要
•Semi-supervised semantic segmentation based on an ensemble of deep learning models.•All models are trained jointly in a co-training setting.•Coherence among models is enforced by minimizing the Jensen-Shannon divergence of the probabilities distributions on the unlabeled samples.•Diversity among models is preserved by enforcing similarity between the prediction of a model on an unlabeled sample and the adversarial prediction of another model on the same sample.•Results on ACDC dataset and SCGM dataset show the capability of our model to outperform previous ensemble approaches by a significant margin.
论文关键词：Deep learning,Semi-supervised learning,Ensemble learning,Co-training,Image segmentation
论文评审过程：Received 16 March 2019, Revised 16 January 2020, Accepted 10 February 2020, Available online 15 February 2020, Version of Record 20 June 2020.
论文官网地址：https://doi.org/10.1016/j.patcog.2020.107269
