mlsanityData to 14 Oct 2026

Encoder-Only Image Registration

arXiv 2509.00451 ↗First posted 30 Aug 2025cs.CVcited by 3

Not on the map. Why?

Abstract (arXiv)

Learning-based techniques have significantly improved the accuracy and speed of deformable image registration. However, challenges such as reducing computational complexity and handling large deformations persist. To address these challenges, we analyze how convolutional neural networks (ConvNets) influence registration performance using the Horn-Schunck optical flow equation. Supported by prior studies and our empirical experiments, we observe that ConvNets play two key roles in registration: linearizing local intensities and harmonizing global contrast variations. Based on these insights, we propose the Encoder-Only Image Registration (EOIR) framework, designed to achieve a better accuracy-efficiency trade-off. EOIR separates feature learning from flow estimation, employing only a 3-layer ConvNet for feature extraction and a set of 3-layer flow estimators to construct a Laplacian feature pyramid, progressively composing diffeomorphic deformations under a large-deformation model. Results on five datasets across different modalities and anatomical regions demonstrate EOIR's effectiveness, achieving superior accuracy-efficiency and accuracy-smoothness trade-offs. With comparable accuracy, EOIR provides better efficiency and smoothness, and vice versa. The source code of EOIR is publicly available on https://github.com/XiangChen1994/EOIR.

Key facts. Cited by 3 papers in the map (data through 14 Oct 2026). Builds on 55 of its 99 references that are in the map. First posted on arXiv 30 Aug 2025.

Builds on

55 of 99 references.

References · 55

  1. 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 30,186
  2. 2015U-Net: Convolutional Networks for Biomedical Image Segmentationcited by 19,713
  3. 2021Swin Transformer: Hierarchical Vision Transformer using Shifted Windowscited by 12,139
  4. 2018MobileNetV2: Inverted Residuals and Linear Bottleneckscited by 4,795
  5. 2016Instance Normalization: The Missing Ingredient for Fast Stylizationcited by 1,198
  6. 2018VoxelMorph: A Learning Framework for Deformable Medical Image Registrationcited by 662
  7. 2017Understanding the Effective Receptive Field in Deep Convolutional Neural Networkscited by 553
  8. 2019A large annotated medical image dataset for the development and evaluation of segmentation algorithmscited by 373
  9. 2014Deeply-Supervised Netscited by 367
  10. 2018Boundary loss for highly unbalanced segmentationcited by 190
  11. 2019Unsupervised Learning of Probabilistic Diffeomorphic Registration for Images and Surfacescited by 186
  12. 2020How Much Position Information Do Convolutional Neural Networks Encode?cited by 172
38 more
  1. 2018An Unsupervised Learning Model for Deformable Medical Image Registrationcited by 150
  2. 2020SynthMorph: learning contrast-invariant registration without acquired imagescited by 140
  3. 2020Large Deformation Diffeomorphic Image Registration with Laplacian Pyramid Networkscited by 136
  4. 2021Learn2Reg: comprehensive multi-task medical image registration challenge, dataset and evaluation in the era of deep learningcited by 125
  5. 2018Unsupervised Learning for Fast Probabilistic Diffeomorphic Registrationcited by 116
  6. 2020Fast Symmetric Diffeomorphic Image Registration with Convolutional Neural Networkscited by 107
  7. 2019Recursive Cascaded Networks for Unsupervised Medical Image Registrationcited by 105
  8. 2021HyperMorph: Amortized Hyperparameter Learning for Image Registrationcited by 85
  9. 2021Topology-Aware Segmentation Using Discrete Morse Theorycited by 75
  10. 2019Unsupervised 3D End-to-End Medical Image Registration with Volume Tweening Networkcited by 62
  11. 2024uniGradICON: A Foundation Model for Medical Image Registrationcited by 58
  12. 2019Dual-Stream Pyramid Registration Networkcited by 52
  13. 2022XMorpher: Full Transformer for Deformable Medical Image Registration via Cross Attentioncited by 46
  14. 2021Fast 3D registration with accurate optimisation and little learning for Learn2Reg 2021cited by 40
  15. 2022U-Net vs Transformer: Is U-Net Outdated in Medical Image Registration?cited by 40
  16. 2024Correlation-aware Coarse-to-fine MLPs for Deformable Medical Image Registrationcited by 38
  17. 2020DeepFLASH: An Efficient Network for Learning-based Medical Image Registrationcited by 37
  18. 2021Learning a Model-Driven Variational Network for Deformable Image Registrationcited by 25
  19. 2023A Robust and Interpretable Deep Learning Framework for Multi-modal Registration via Keypointscited by 25
  20. 2022On Finite Difference Jacobian Computation in Deformable Image Registrationcited by 23
  21. 2022Fourier-Net: Fast Image Registration with Band-limited Deformationcited by 22
  22. 2019One Shot Learning for Deformable Medical Image Registration and Periodic Motion Trackingcited by 21
  23. 2023Non-iterative Coarse-to-fine Transformer Networks for Joint Affine and Deformable Image Registrationcited by 21
  24. 2022Diffeomorphic Image Registration with Neural Velocity Fieldcited by 20
  25. 2022Medical Image Registration via Neural Fieldscited by 19
  26. 2023Spatially Covariant Image Registration with Text Promptscited by 18
  27. 2021A Deep Discontinuity-Preserving Image Registration Networkcited by 17
  28. 2023Improving Misaligned Multi-modality Image Fusion with One-stage Progressive Dense Registrationcited by 17
  29. 2023SAMConvex: Fast Discrete Optimization for CT Registration using Self-supervised Anatomical Embedding and Correlation Pyramidcited by 14
  30. 2023Heteroscedastic Uncertainty Estimation Framework for Unsupervised Registrationcited by 14
  31. 2024MemWarp: Discontinuity-Preserving Cardiac Registration with Memorized Anatomical Filterscited by 13
  32. 2025Beyond the LUMIR challenge: The pathway to foundational registration modelscited by 12
  33. 2023Spatially Covariant Lesion Segmentationcited by 10
  34. 2022Learning correspondences of cardiac motion from images using biomechanics-informed modelingcited by 8
  35. 2023Towards an accurate and generalizable multiple sclerosis lesion segmentation model using self-ensembled lesion fusioncited by 6
  36. 2024Slicer Networkscited by 6
  37. 2024UltraCortex: Submillimeter Ultra-High Field 9.4 T Brain MR Image Collection and Manual Cortical Segmentationscited by 6
  38. 2024Bi-Directional MS Lesion Filling and Synthesis Using Denoising Diffusion Implicit Model-based Lesion Repaintingcited by 6

50 most cited of 55 All 55 (JSON)

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 1; 2026: 2.

3 papers.

  1. 2025MSRepaint: Multiple Sclerosis Repaint with Conditional Denoising Diffusion Implicit Model for Bidirectional Lesion Filling and Synthesiscited by 1
  2. 2026FMIR, a foundation model-based Image Registration Framework for Robust Image Registrationcited by 0
  3. 2026Reg-TTR, Test-Time Refinement for Fast, Robust and Accurate Image Registrationcited by 0