mlsanityData to 7 Oct 2026

vMFNet: Compositionality Meets Domain-generalised Segmentation

arXiv 2206.14538 ↗First posted 29 Jun 2022cs.CVcited by 11

Not on the map. Why?

Abstract (arXiv)

Training medical image segmentation models usually requires a large amount of labeled data. By contrast, humans can quickly learn to accurately recognise anatomy of interest from medical (e.g. MRI and CT) images with some limited guidance. Such recognition ability can easily generalise to new images from different clinical centres. This rapid and generalisable learning ability is mostly due to the compositional structure of image patterns in the human brain, which is less incorporated in medical image segmentation. In this paper, we model the compositional components (i.e. patterns) of human anatomy as learnable von-Mises-Fisher (vMF) kernels, which are robust to images collected from different domains (e.g. clinical centres). The image features can be decomposed to (or composed by) the components with the composing operations, i.e. the vMF likelihoods. The vMF likelihoods tell how likely each anatomical part is at each position of the image. Hence, the segmentation mask can be predicted based on the vMF likelihoods. Moreover, with a reconstruction module, unlabeled data can also be used to learn the vMF kernels and likelihoods by recombining them to reconstruct the input image. Extensive experiments show that the proposed vMFNet achieves improved generalisation performance on two benchmarks, especially when annotations are limited. Code is publicly available at: https://github.com/vios-s/vMFNet.

Key facts. Cited by 11 papers in the map (data through 7 Oct 2026). Builds on 29 of its 43 references that are in the map. First posted on arXiv 29 Jun 2022.

Builds on

29 of 43 references.

References · 29

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2014Adam: A Method for Stochastic Optimizationcited by 36,960
  3. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,160
  4. 2017Prototypical Networks for Few-shot Learningcited by 3,001
  5. 2019Domain Generalization by Solving Jigsaw Puzzlescited by 449
  6. 2019Domain Generalization via Model-Agnostic Learning of Semantic Featurescited by 405
  7. 2021FedDG: Federated Domain Generalization on Medical Image Segmentation via Episodic Learning in Continuous Frequency Spacecited by 284
  8. 2017Emergence of Invariance and Disentanglement in Deep Representationscited by 213
  9. 2021FSDR: Frequency Space Domain Randomization for Domain Generalizationcited by 163
  10. 2019Deep learning for cardiac image segmentation: A reviewcited by 131
  11. 2020Shape-aware Meta-learning for Generalizing Prostate MRI Segmentation to Unseen Domainscited by 124
  12. 2020Domain Generalization for Medical Imaging Classification with Linear-Dependency Regularizationcited by 122
17 more
  1. 2020Test-Time Adaptable Neural Networks for Robust Medical Image Segmentationcited by 102
  2. 2021Learning to Compose Visual Relationscited by 72
  3. 2019Disentangled Representation Learning in Cardiac Image Analysiscited by 69
  4. 2022Enhancing Pseudo Label Quality for Semi-Supervised Domain-Generalized Medical Image Segmentationcited by 51
  5. 2018Learning Compositional Representations for Few-Shot Recognitioncited by 50
  6. 2020Compositional Convolutional Neural Networks: A Deep Architecture with Innate Robustness to Partial Occlusioncited by 47
  7. 2021Semi-supervised Meta-learning with Disentanglement for Domain-generalised Medical Image Segmentationcited by 36
  8. 2021Compositional Transformers for Scene Generationcited by 24
  9. 2020Robust Instance Segmentation through Reasoning about Multi-Object Occlusioncited by 23
  10. 2021AtrialGeneral: Domain Generalization for Left Atrial Segmentation of Multi-Center LGE MRIscited by 16
  11. 2021Domain Composition and Attention for Unseen-Domain Generalizable Medical Image Segmentationcited by 16
  12. 2020Disentangled Representations for Domain-generalized Cardiac Segmentationcited by 14
  13. 2020Measuring the Biases and Effectiveness of Content-Style Disentanglementcited by 14
  14. 2021Cooperative Training and Latent Space Data Augmentation for Robust Medical Image Segmentationcited by 9
  15. 2021Anatomy of Domain Shift Impact on U-Net Layers in MRI Segmentationcited by 9
  16. 2021Re-using Adversarial Mask Discriminators for Test-time Training under Distribution Shiftscited by 7
  17. 2021Stop Throwing Away Discriminators! Re-using Adversaries for Test-Time Trainingcited by 4

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2023: 3; 2024: 3; 2025: 4; 2026: 1.

11 papers.

  1. 2023Towards Generic Semi-Supervised Framework for Volumetric Medical Image Segmentationcited by 35
  2. 2024RaffeSDG: Random Frequency Filtering enabled Single-source Domain Generalization for Medical Image Segmentationcited by 3
  3. 2023Compositionally Equivariant Representation Learningcited by 2
  4. 2025FedSemiDG: Domain Generalized Federated Semi-supervised Medical Image Segmentationcited by 2
  5. 2024S&D Messenger: Exchanging Semantic and Domain Knowledge for Generic Semi-Supervised Medical Image Segmentationcited by 1
  6. 2025Dual-supervised Asymmetric Co-training for Semi-supervised Medical Domain Generalizationcited by 1
  7. 2023Compositional Representation Learning for Brain Tumour Segmentationcited by 0
  8. 2024Enhancing Cross-Modal Medical Image Segmentation through Compositionalitycited by 0
  9. 2025Casual Inference via Style Bias Deconfounding for Domain Generalizationcited by 0
  10. 2025Diverse Teaching and Label Propagation for Generic Semi-Supervised Medical Image Segmentationcited by 0
  11. 2026SNR-Adaptive Unified Diffusion for Multi-Task Medical Image Segmentationcited by 0