Effect of latent space distribution on the segmentation of images with multiple annotations

Ishaan Bhat, Josien P. W. Pluim, Max A. Viergever, Hugo J. Kuijf
4/26/2023
cs.CVcs.LGeess.IV

Official code: ishaanb92/GeneralizedProbabilisticUNet (16 stars, MIT, last commit 2024-03-13).

Abstract

We propose the Generalized Probabilistic U-Net, which extends the Probabilistic U-Net by allowing more general forms of the Gaussian distribution as the latent space distribution that can better approximate the uncertainty in the reference segmentations. We study the effect the choice of latent space distribution has on capturing the variation in the reference segmentations for lung tumors and white matter hyperintensities in the brain. We show that the choice of distribution affects the sample diversity of the predictions and their overlap with respect to the reference segmentations. We have made our implementation available at https://github.com/ishaanb92/GeneralizedProbabilisticUNet

View on arXivView PDF

Code Implementations(1)

ishaanb92/GeneralizedProbabilisticUNetOfficial100%
MIT

PyTorch model for the Generalized Probabilistic U-Net. For more details see: https://www.melba-journal.org/papers/2023:005.html

169PythonMay 19, 20222 years agoMIT

Similar Papers

A Probabilistic Segment Anything Model for Ambiguity-Aware Medical Image Segmentation

84% match1 repo52025

SegReg: Latent Space Regularization for Improved Medical Image Segmentation

82% match2026

Variance-Penalized MC-Dropout as a Learned Smoothing Prior for Brain Tumour Segmentation

79% match2026

Probabilistic Modeling of Inter- and Intra-observer Variability in Medical Image Segmentation

78% match2023

Probabilistic Modeling of Multi-rater Medical Image Segmentation for Diversity and Personalization

78% match1 repo72025

Semi-Supervised Generative Learning via Latent Space Distribution Matching

78% match2026

Cite this paper

@article{bhat2023effect,
  title  = {Effect of latent space distribution on the segmentation of images with multiple annotations},
  author = {Ishaan Bhat and Josien P. W. Pluim and Max A. Viergever and Hugo J. Kuijf},
  year   = {2023},
  eprint = {2304.13476},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2304.13476v1}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2304.13476.svg)](https://codeofpaper.com/paper/2304.13476)

Discussion