mlsanityData to 14 Oct 2026

Noisy Image Segmentation With Soft-Dice

arXiv 2304.00801 ↗First posted 3 Apr 2023cs.CVcited by 1

Not on the map. Why?

Abstract (arXiv)

This paper presents a study on the soft-Dice loss, one of the most popular loss functions in medical image segmentation, for situations where noise is present in target labels. In particular, the set of optimal solutions are characterized and sharp bounds on the volume bias of these solutions are provided. It is further shown that a sequence of soft segmentations converging to optimal soft-Dice also converges to optimal Dice when converted to hard segmentations using thresholding. This is an important result because soft-Dice is often used as a proxy for maximizing the Dice metric. Finally, experiments confirming the theoretical results are provided.

Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Builds on 21 of its 34 references that are in the map. First posted on arXiv 3 Apr 2023.

Builds on

21 of 34 references.

References · 21

  1. 2015U-Net: Convolutional Networks for Biomedical Image Segmentationcited by 19,713
  2. 2017Generalised Dice overlap as a deep learning loss function for highly unbalanced segmentationscited by 620
  3. 2019Embracing Imperfect Datasets: A Review of Deep Learning Solutions for Medical Image Segmentationcited by 260
  4. 2019Deep learning with noisy labels: exploring techniques and remedies in medical image analysiscited by 173
  5. 2020U-Net and its variants for medical image segmentation: theory and applicationscited by 169
  6. 2016The Importance of Skip Connections in Biomedical Image Segmentationcited by 126
  7. 2019Optimizing the Dice Score and Jaccard Index for Medical Image Segmentation: Theory & Practicecited by 71
  8. 2020Optimization for Medical Image Segmentation: Theory and Practice when evaluating with Dice Score or Jaccard Indexcited by 52
  9. 2020Deep Convolutional Neural Networks with Spatial Regularization, Volume and Star-shape Priori for Image Segmentationcited by 31
  10. 2020SoftSeg: Advantages of soft versus binary training for image segmentationcited by 26
  11. 2020Post Training Uncertainty Calibration of Deep Networks For Medical Image Segmentationcited by 19
  12. 2021Using Soft Labels to Model Uncertainty in Medical Image Segmentationcited by 18
9 more
  1. 2018Imperfect Segmentation Labels: How Much Do They Matter?cited by 14
  2. 2022Label fusion and training methods for reliable representation of inter-rater uncertaintycited by 13
  3. 2019Soft labeling by Distilling Anatomical knowledge for Improved MS Lesion Segmentationcited by 11
  4. 2021On the relationship between calibrated predictors and unbiased volume estimationcited by 11
  5. 2019Calibrated Surrogate Maximization of Linear-fractional Utility in Binary Classificationcited by 10
  6. 2021Label noise in segmentation networks : mitigation must deal with biascited by 10
  7. 2020Superpixel-Guided Label Softening for Medical Image Segmentationcited by 6
  8. 2022On Image Segmentation With Noisy Labels: Characterization and Volume Properties of the Optimal Solutions to Accuracy and Dicecited by 6
  9. 2022Theoretical analysis and experimental validation of volume bias of soft Dice optimized segmentation maps in the context of inherent uncertaintycited by 6

Built on by

Cited by year

Citing papers per year: 2023: 1.

1 paper.

  1. 2023Marginal Thresholding in Noisy Image Segmentationcited by 0