Bayesian Pseudo Labels: Expectation Maximization for Robust and Efficient Semi-Supervised Segmentation.
Mou-Cheng XuYukun ZhouChen JinMarius de GrootDaniel C. AlexanderNeil P. OxtobyYipeng HuJoseph JacobPublished in: CoRR (2022)
Keyphrases
- semi supervised
- expectation maximization
- robust segmentation
- em algorithm
- image labeling
- image segmentation
- segmentation algorithm
- pairwise
- computationally efficient
- training set
- multiscale
- segmentation method
- medical images
- level set
- fully supervised
- maximum likelihood
- labeled data
- energy function
- segmented images
- labeling process
- label noise
- minimum description length
- fully unsupervised
- label propagation
- generative model
- text classification
- training data