Pith. sign in

REVIEW 4 major objections 4 minor 3 cited by

Invariant Shape Representation Learning For Image Classification

T0 review · 4 major / 4 minor · reviewed 2026-08-12 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Joint shape-invariance learning outperforms image-only models.

desk verdict Sensible integration of shape features and IRM, but real-data evaluation doesn't test the cross-environment claim. read the letter →

arxiv 2411.12201 v1 pith:AAOZQZAU submitted 2024-11-19 cs.CV

Tonmoy Hossain , Jing Ma , Jundong Li , Miaomiao Zhang This is my paper · ORCID
classification cs.CV
keywords invariantriskminimizationshaperepresentationlearningdeformation-basedanalysisimageclassificationdomaingeneralizationspuriouscorrelationmedicalgeodesicshooting
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper tries to establish that image classifiers become more accurate and more stable across changed environments when they learn invariant features in a joint space of raw images and deformation-based shape representations, rather than in image space alone. It proposes ISRL, which combines a template-registration network that encodes each image as an initial velocity field (the shape code) with an invariant risk minimization (IRM) penalty that forces the fused image+shape predictor to be simultaneously good in every training environment. The claim is supported by experiments on simulated 2D drawings with color as a spurious factor, on 3D brain MRIs where age is the confounder, and on cardiac MRI videos where heart-failure history is the confounder; in each setting ISRL reports higher accuracy and F1 than ERM, image-only IRM, and four domain-generalization baselines.

What carries the argument

The central object is the joint loss $\ell_{\mathrm{ISRL}} = \ell_{\mathrm{SRL}} + \beta \ell_{\mathrm{IC}}$, where $\ell_{\mathrm{SRL}}$ trains an encoder-decoder to predict the initial velocity field $v_n(0)$ of a geodesic deformation, solved via the Euler–Poincaré differential equation, that maps a template to each input image, and $\ell_{\mathrm{IC}}$ is the invariant classification loss built from the IRM penalty $\sum_{e} R_e(\psi) + \lambda \|\nabla_{w=1.0} R_e(w \cdot \psi)\|_2^2$. The shape code $v$ and image code $I$ are concatenated in a latent space and fed to the classifier. The mechanism that carries the argument is the pressure the IRM gradient penalty puts on the fused representation to be optimal in every environment simultaneously, which is what supposedly removes spurious factors while preserving the shape features that genuinely determine the label.

What would settle it

Train ISRL and image-only IRM on the same three datasets with deliberately misspecified environments, for example random age splits for ADNI and random CHF splits for the cardiac data. If ISRL's advantage over IRM persists under misspecified environments, then the invariance mechanism is not what produces the gains; if the advantage disappears, the result is confirmed to depend on the environments containing the true spurious correlations.

Watch

Extended reading notes

Core claim

The paper's central claim is that shape features, when learned jointly with image features under an invariant risk minimization penalty, contain the invariant information needed to suppress spurious correlations that fool ordinary classifiers. The core discovery is the framework itself: a geometric shape learning network (an encoder-decoder that predicts the initial velocity field of a geodesic deformation aligning an image to a template) feeding into an invariant classification network, whose loss adds the environment-wise IRM gradient penalty to the shape reconstruction loss. With this coupling, the model consistently outperforms image-only invariant learning and empirical risk minimization on all tested backbones: on 2D Quickdraw with color confounds, ISRL reaches 87.56% accuracy under ResNet versus 71.11% for IRM; on 3D brain MRIs it reaches 84.44% versus 72.67%; on cardiac videos 88.76% versus 86.51% under ViViT. The paper argues these gains come from leveraging shape's invariance to intensity and texture changes while IRM discards environment-specific spurious shape changes.

Load-bearing premise

The load-bearing premise is that the chosen environments—color on Quickdraw, age on brain MRIs, and congestive heart failure on cardiac MRIs—are exactly where the spurious correlations live, and that the training environments cover the test distribution, so the IRM penalty can identify a single predictor that is optimal across all of them.

Editorial extensions

If this is right

  • ISRL reports gains across four backbones on 2D and 3D tasks, so the combination of shape features and IRM is not tied to a single network architecture.
  • On the 2D Quickdraw task, raising the label-flip probability from 0% to 50% causes the baselines to drop 30–55% in accuracy while ISRL drops only about 13–20%, suggesting the learned representation retains useful shape cues even when the spurious color-label link weakens.
  • The joint training scheme beats the disjoint two-step pipeline by 2–3%, so optimizing shape reconstruction and invariant classification together, rather than as a pre-processing stage, contributes to the reported performance.
  • The ablation shows that adding geometric shape features improves both ERM and IRM, and adding IRM improves both image-only and image+shape models; ISRL, which does both, is best, indicating the two ingredients are additive.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Inference: If the reported gains hold under properly validated environments, the same joint shape-invariance recipe could be applied to other medical imaging tasks with known confounders, such as imaging-site batch effects, sex, or medication history, where shape codes are already available from registration pipelines.
  • Inference: The paper's environment definitions for the medical datasets are coarse and potentially miscalibrated (for example, each age environment is described as 150 images while the total is 690, and the 65/25/10 split is unexplained), so a careful reader should re-run with well-defined, non-overlapping environments before treating the medical numbers as decisive evidence.
  • Inference: Because the IRM penalty is sensitive to the chosen environments and penalty weight, ISRL's advantage could partly come from the extra shape reconstruction signal acting as a regularizer rather than from true invariance; an isolation test would decide between the causal and the regularization explanation.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 4 minor

Summary. The paper proposes ISRL, a framework that combines deformation-based geometric shape features with image features and trains a classifier under an invariant risk minimization (IRM) penalty to improve robustness across environments. The method is evaluated on a synthetic 2D colored Quickdraw benchmark, 3D brain MRIs (ADNI) with age as environment, and 3D cardiac MRI videos with congestive heart failure as environment, comparing against ERM, IRM, and several domain generalization baselines across multiple backbones. The authors report consistent accuracy gains, especially on the synthetic benchmark, and provide code.

Significance. If the claims hold, the paper would make a useful contribution by extending invariant representation learning to an integrated image-plus-deformation shape space, an area that has not been explored. The synthetic benchmark with a held-out test environment is a strength, as is the public code release and the inclusion of diverse backbones and medical imaging tasks. However, the real-data experiments do not currently validate the central cross-environment generalization claim, since the test splits are random splits of the same environment pools rather than held-out environments. The reported numbers also lack any measure of variance, which weakens the claim of consistent improvement. The significance is therefore conditional on substantially strengthening the experimental protocol.

major comments (4)
  1. [§4.2, Tables 2 and 3; §4 dataset descriptions] The central claim of cross-environment robustness is not tested on the real 3D datasets. For brain MRI, the three age environments are all included in the training split, and the 65/25/10 split appears to be a random split of the full 690-image pool; for cardiac MRI, the two CHF environments are both in training, with a random 70/15/15 split. In neither case is a model evaluated on an unseen age group or an unseen CHF status. A random split within the same environments does not measure robustness to distribution shift, because environment-specific shortcuts present in training also appear in the test split. To support the stated motivation, the authors should either evaluate on held-out environments (e.g., train on two age groups and test on the third, or train only on non-CHF subjects and test on CHF subjects) or explicitly report per-environment test accuracy and frame the real-data results as same-distribution feature learning rather than invariance. As written, the paper's own Limitation section concedes that IRM can fail when training distributions do not cover test scenarios, but the current protocol does not even expose the model to that scenario.
  2. [§4, 2D simulated data paragraph] The synthetic test protocol is ambiguous. The text states that three environments are created with pe = 0.2, 0.1, 0.9 for two training and one testing environment, but then says the synthesized dataset is split 70% train, 15% validation, 15% test. If the 15% test split is a random subset of all three environments, then the test set includes examples from the training environments and is not a held-out environment. If the test set is restricted to the pe = 0.9 environment, the split description and the reported percentages are misleading. Please clarify how the split is applied, and ideally report results with the test set consisting solely of the pe = 0.9 environment, since that is the only direct evidence for cross-environment generalization in the paper.
  3. [§4, 3D brain MRI paragraph] The ADNI environment definition is internally inconsistent. The text says there are three age environments (50s–60s, 70s, 80s–90s) and that each environment includes a total of 150 images, which would be 450 images in total, yet the dataset is stated to contain 690 images. The 65/25/10 split of 690 images yields non-integer counts (448.5, 172.5, 69), so it is unclear how the split is applied and whether the environments are balanced and disjoint. Please report the exact number of subjects per environment, the number per class per environment, and how the split is performed (per environment or globally).
  4. [§4.2, Tables 1–3 and Fig. 3] No measure of variance is reported for any result. All accuracy, precision, and F1 values are point estimates from a single run or a single split. Several reported gains are modest (e.g., 88.76 vs. 86.51 for ViViT on cardiac MRI, and 76.67 vs. 74.00 for CNN on brain MRI), and without repeated runs it is impossible to determine whether the differences are statistically meaningful. The paper's phrasing 'consistently outperforms' requires supporting evidence such as means and standard deviations over multiple seeds, or a paired significance test. This is particularly important for the real-data results, where the differences are smaller than on the synthetic benchmark.
minor comments (4)
  1. [Figure 1 caption] The figure contains a typo: 'Geomtric' should be 'Geometric'.
  2. [Fig. 3 caption] The sentence 'Overall, our model consistently archives higher accuracy (> 10%) than the baselines' should read 'achieves'.
  3. [§3.2, Eq. (5)] The notation in Eq. (5) is difficult to parse, with a scalar w = 1.0 described as a 'dummy classifier'. Please rewrite this equation more clearly, specifying exactly which variables are held fixed when the gradient is computed, since the penultimate sentence of §3.2 says 'Minimizing the loss above' but the formal statement of the optimization objective is ambiguous.
  4. [§4, cardiac MRI paragraph] The sentence 'we found 146 videos have CHF (E1, 48/98 as scar/non-scar) and 364 not (E2, 210/154 as scar/non-scar)' could be clarified by stating the total scar/non-scar counts (which appear to be 258 scar and 252 non-scar) so that the class balance across environments is immediately interpretable.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the ISRL objective is the standard IRMv1 penalty applied to an independently published shape-feature backbone; reported gains are empirical, not derived from fitted parameters or self-referential definitions.

full rationale

The paper's derivation chain is self-contained rather than circular. The geometric shape representation loss (Eq. 4) is the deformation-based shape learning objective taken from Geo-SIC, an independently peer-reviewed NeurIPS paper with public code, and the invariant classification loss (Eq. 5) is the standard IRMv1 penalty from Arjovsky et al. The novelty is the joint combination of these two independent components via alternating optimization, and no result is derived by defining an input in terms of an output. Hyperparameter lambda is selected by cross-validation for both ISRL and the IRM baseline, so the reported accuracies are measured on held-out test splits rather than being the fitted objective itself. The 2D synthetic experiment uses a held-out test environment (pe=0.9), providing an external check of invariance. Self-citations to Geo-SIC and other same-group works are to independently published, peer-reviewed methods used as building blocks, not to forbidden alternatives or to a uniqueness theorem that forces the paper's choice. The paper's own Limitation section candidly states conditions under which IRM can fail, and the noted inconsistencies in the ADNI environment description and the within-environment random splits for 3D data are experimental-design/validation weaknesses, not circular reasoning. No equation reduces to its own input, and no fitted parameter is renamed as a prediction. Therefore, the appropriate circularity score is 0.

Assumptions & free parameters 5 free parameters · 4 assumptions · 0 invented entities

The central claim rests on existing, independently published machinery: LDDMM deformation features (Geo-SIC) and the IRMv1 penalty. The genuinely new contribution, the joint integration, depends on several hand-chosen hyperparameters (lambda, beta, alpha, sigma, integration steps) and on domain assumptions about which environments contain the spurious correlations. No fundamentally new entity or constant is introduced.

free parameters (5)
  • lambda (invariance penalty weight) = 10000 to 100000
    Tuned per dataset via cross-validation; the reported accuracies in Table 4 depend on the chosen lambda.
  • beta (joint loss weight) = not reported
    Weight balancing shape learning and invariant classification losses; value is required to reproduce the method but omitted from the paper.
  • alpha (regularity in operator L) = 3
    Fixed parameter in the differential operator L = -alpha * Delta + I for the deformation metric.
  • sigma (noise variance) = 0.02
    Used in the shape representation loss in Eq (4).
  • time steps for EPDiff integration = 10
    Euler integration steps for geodesic shooting in Eq (2), a numerical approximation choice.
assumptions (4)
  • domain assumption Deformable template model: images are deformed versions of a common template and the deformation encodes shape.
    Section 2, underlying the entire shape representation; standard in LDDMM but not proved for the specific datasets.
  • standard math EPDiff and geodesic shooting (Eq 2) uniquely determine geodesics.
    Section 2, from Vialard et al. and Miller et al.; accepted background in computational anatomy.
  • domain assumption IRM assumption: there exists a predictor w that is simultaneously optimal for all environments.
    Definition 3.1 and Section 3.2; the paper's Limitation section acknowledges this can fail in practice.
  • domain assumption The chosen environments (color, age, CHF) capture the relevant spurious correlations.
    Section 4; if age or CHF are not the active confounders, the invariant features learned may not transfer to unseen environments.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Invariant Shape Representation Learning For Image Classification." pith.science (2026). https://pith.science/paper/AAOZQZAU

@misc{pith2026241112201,
  author       = {Pith},
  title        = {Pith review of: Invariant Shape Representation Learning For Image Classification},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/AAOZQZAU}},
  note         = {Machine review of arXiv:2411.12201}
}
read the original abstract

Geometric shape features have been widely used as strong predictors for image classification. Nevertheless, most existing classifiers such as deep neural networks (DNNs) directly leverage the statistical correlations between these shape features and target variables. However, these correlations can often be spurious and unstable across different environments (e.g., in different age groups, certain types of brain changes have unstable relations with neurodegenerative disease); hence leading to biased or inaccurate predictions. In this paper, we introduce a novel framework that for the first time develops invariant shape representation learning (ISRL) to further strengthen the robustness of image classifiers. In contrast to existing approaches that mainly derive features in the image space, our model ISRL is designed to jointly capture invariant features in latent shape spaces parameterized by deformable transformations. To achieve this goal, we develop a new learning paradigm based on invariant risk minimization (IRM) to learn invariant representations of image and shape features across multiple training distributions/environments. By embedding the features that are invariant with regard to target variables in different environments, our model consistently offers more accurate predictions. We validate our method by performing classification tasks on both simulated 2D images, real 3D brain and cine cardiovascular magnetic resonance images (MRIs). Our code is publicly available at https://github.com/tonmoy-hossain/ISRL.

Figures

Figures reproduced from arXiv: 2411.12201 by the authors.

Figure 1
Figure 1. An overview of our proposed network architecture of ISRL. The geometric shape learning ( [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Top to bottom: examples of 2D simulated data vs. 3D [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. A comparison of the baselines and ISRL on four differ [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Comparison of disjoint (two-step) learning vs. our joint [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: A visual comparison of activation maps generated by [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 3 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Learning To Focus: Anatomy-Guided Attention Regularization for Medical Image Classification

    cs.CV 2026-07 conditional novelty 6.0 of 10

    A hinge regularizer on foundation-model anatomical masks forces classifiers to keep more attention energy on foreground than background, improving accuracy and localization without pixel labels.

  2. ShapeEmbed: a self-supervised learning framework for 2D contour quantification

    cs.CV 2025-07 conditional novelty 6.0 of 10

    A VAE trained on normalized distance matrices of object contours yields shape descriptors invariant to similarity transforms and point re-indexing, and these descriptors improve downstream shape classification.

  3. Learning to Unify Deformable Shape and Texture Representations for Cardiac Video Classification

    cs.CV 2026-07 conditional novelty 5.0 of 10

    ShapeFuse uses bidirectional cross-modal temporal attention and adaptive gating to fuse deformable shape and texture features for cardiac video classification, outperforming existing fusion strategies on a cine CMR dataset.

Reference graph

Works this paper leans on

69 extracted references · 56 canonical work pages · cited by 3 Pith papers

  1. [1]

    Invariant risk minimization games

    Kartik Ahuja, Karthikeyan Shanmugam, Kush Varshney, and Amit Dhurandhar. Invariant risk minimization games. In International Conference on Machine Learning, pages 145–

    work page

  2. [2]

    Invariant risk minimization

    Martin Arjovsky, L ´eon Bottou, Ishaan Gulrajani, and David Lopez-Paz. Invariant risk minimization. arXiv preprint arXiv:1907.02893, 2019. 1, 3, 4, 5, 6, 8

    arXiv 1907

  3. [3]

    Vivit: A video vision transformer

    Anurag Arnab, Mostafa Dehghani, Georg Heigold, Chen Sun, Mario Lu ˇci´c, and Cordelia Schmid. Vivit: A video vision transformer. In Proceedings of the IEEE/CVF inter- national conference on computer vision , pages 6836–6846,

  4. [4]

    Sur la g ´eom´etrie diff ´erentielle des groupes de lie de dimension infinie et ses applications `a l’hydrodynamique des fluides parfaits

    Vladimir Arnold. Sur la g ´eom´etrie diff ´erentielle des groupes de lie de dimension infinie et ses applications `a l’hydrodynamique des fluides parfaits. In Annales de l’institut Fourier, volume 16, pages 319–361, 1966. 2

    work page 1966

  5. [5]

    Computing large deformation metric map- pings via geodesic flows of diffeomorphisms

    MIRZA Faisal Beg, Michael I Miller, Alain Trouv ´e, and Laurent Younes. Computing large deformation metric map- pings via geodesic flows of diffeomorphisms. International journal of computer vision, 61(2):139–157, 2005. 2

    work page 2005

  6. [6]

    Learning distributions of shape trajectories from longitudinal datasets: a hierarchical model on a manifold of diffeomor- phisms

    Alexandre B ˆone, Olivier Colliot, and Stanley Durrleman. Learning distributions of shape trajectories from longitudinal datasets: a hierarchical model on a manifold of diffeomor- phisms. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 9271–9280, 2018. 2

    work page 2018

  7. [7]

    Parametriza- tion of closed surfaces for 3-d shape description

    Ch Brechb ¨uhler, Guido Gerig, and Olaf K¨ubler. Parametriza- tion of closed surfaces for 3-d shape description. Computer vision and image understanding, 61(2):154–170, 1995. 2

    work page 1995

  8. [8]

    Causal- ity matters in medical imaging

    Daniel C Castro, Ian Walker, and Ben Glocker. Causal- ity matters in medical imaging. Nature Communications, 11(1):3673, 2020. 1

    work page 2020

Show all 69 references
  1. [9]

    Semi-supervised task-driven data augmentation for medical image segmentation

    Krishna Chaitanya, Neerav Karani, Christian F Baumgart- ner, Ertunc Erdil, Anton Becker, Olivio Donati, and Ender Konukoglu. Semi-supervised task-driven data augmentation for medical image segmentation. Medical Image Analysis , 68:101934, 2021. 4

    2021

  2. [10]

    In- variant rationalization

    Shiyu Chang, Yang Zhang, Mo Yu, and Tommi Jaakkola. In- variant rationalization. In International Conference on Ma- chine Learning, pages 1448–1458. PMLR, 2020. 1, 8

    2020

  3. [11]

    Enhancing mr image segmentation with re- alistic adversarial data augmentation

    Chen Chen, Chen Qin, Cheng Ouyang, Zeju Li, Shuo Wang, Huaqi Qiu, Liang Chen, Giacomo Tarroni, Wenjia Bai, and Daniel Rueckert. Enhancing mr image segmentation with re- alistic adversarial data augmentation. Medical Image Analy- sis, 82:102597, 2022. 4

    2022

  4. [12]

    Transunet: Transformers make strong encoders for medi- cal image segmentation

    Jieneng Chen, Yongyi Lu, Qihang Yu, Xiangde Luo, Ehsan Adeli, Yan Wang, Le Lu, Alan L Yuille, and Yuyin Zhou. Transunet: Transformers make strong encoders for medi- cal image segmentation. arXiv preprint arXiv:2102.04306,

    arXiv

  5. [13]

    Active shape models-their training and application

    Timothy F Cootes, Christopher J Taylor, David H Cooper, and Jim Graham. Active shape models-their training and application. Computer vision and image understanding , 61(1):38–59, 1995. 2

    1995

  6. [14]

    An image is worth 16x16 words: Trans- formers for image recognition at scale

    Alexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weissenborn, Xiaohua Zhai, Thomas Unterthiner, Mostafa Dehghani, Matthias Minderer, Georg Heigold, Syl- vain Gelly, et al. An image is worth 16x16 words: Trans- formers for image recognition at scale. arXiv preprint a...

    2010 arXiv

  7. [15]

    Domain-adversarial training of neural networks

    Yaroslav Ganin, Evgeniya Ustinova, Hana Ajakan, Pas- cal Germain, Hugo Larochelle, Franc ¸ois Laviolette, Mario March, and Victor Lempitsky. Domain-adversarial training of neural networks. Journal of machine learning research , 17(59):1–35, 2016. 2, 5, 7

    2016

  8. [16]

    Imagenet-trained cnns are biased towards texture; increasing shape bias improves accuracy and robustness.arXiv preprint arXiv:1811.12231, 2018

    Robert Geirhos, Patricia Rubisch, Claudio Michaelis, Matthias Bethge, Felix A Wichmann, and Wieland Brendel. Imagenet-trained cnns are biased towards texture; increasing shape bias improves accuracy and robustness.arXiv preprint arXiv:1811.12231, 2018. 1

    2018 arXiv

  9. [17]

    Confidence calibration for domain generalization under covariate shift

    Yunye Gong, Xiao Lin, Yi Yao, Thomas G Dietterich, Ajay Divakaran, and Melinda Gervasio. Confidence calibration for domain generalization under covariate shift. In Proceed- ings of the IEEE/CVF International Conference on Com- puter Vision, pages 8958–8967, 2021. 2

    2021

  10. [18]

    Deep residual learning for image recognition

    Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep residual learning for image recognition. In Proceed- ings of the IEEE conference on computer vision and pattern recognition, pages 770–778, 2016. 3, 5

    2016

  11. [19]

    Diffeo- morphic autoencoders for lddmm atlas building

    Jacob Hinkle, David Womble, and Hong-Jun Yoon. Diffeo- morphic autoencoders for lddmm atlas building. 2018. 3

    2018

  12. [20]

    Fast geodesic regression for population-based image analysis

    Yi Hong, Polina Golland, and Miaomiao Zhang. Fast geodesic regression for population-based image analysis. In International Conference on Medical Image Computing and Computer-Assisted Intervention , pages 317–325. Springer,

  13. [21]

    Brain tumor detection using convolutional neural net- work

    Tonmoy Hossain, Fairuz Shadmani Shishir, Mohsena Ashraf, MD Abdullah Al Nasim, and Faisal Muhammad Shah. Brain tumor detection using convolutional neural net- work. In 2019 1st international conference on advances in science, engineering and robotics technology (ICASERT), page...

    2019

  14. [22]

    Mgaug: Multimodal geometric augmentation in latent spaces of image deforma- tions

    Tonmoy Hossain and Miaomiao Zhang. Mgaug: Multimodal geometric augmentation in latent spaces of image deforma- tions. arXiv preprint arXiv:2312.13440, 2023. 2

    2023 arXiv

  15. [23]

    The alzheimer’s disease neuroimaging initiative (adni): Mri methods

    Clifford R Jack Jr, Matt A Bernstein, Nick C Fox, Paul Thompson, Gene Alexander, Danielle Harvey, Bret Borowski, Paula J Britson, et al. The alzheimer’s disease neuroimaging initiative (adni): Mri methods. Journal of Magnetic Resonance Imaging, 27(4):685–691, 2008. 2, 5

    2008

  16. [24]

    Sadir: shape-aware diffusion models for 3d image recon- struction

    Nivetha Jayakumar, Tonmoy Hossain, and Miaomiao Zhang. Sadir: shape-aware diffusion models for 3d image recon- struction. In International Workshop on Shape in Medical Imaging, pages 287–300. Springer, 2023. 1

    2023

  17. [25]

    The quick, draw!-ai experi- ment

    Jonas Jongejan, Henry Rowley, Takashi Kawashima, Jong- min Kim, and Nick Fox-Gieg. The quick, draw!-ai experi- ment. Mount View, CA, accessed Feb, 17(2018):4, 2016. 2, 4

    2018

  18. [26]

    Improved image registration by sparse patch-based deformation estimation

    Minjeong Kim, Guorong Wu, Qian Wang, Seong-Whan Lee, and Dinggang Shen. Improved image registration by sparse patch-based deformation estimation. Neuroimage, 105:257– 268, 2015. 3

    2015

  19. [27]

    Arrhythmias and mortality in conges- tive heart failure

    John Kjekshus. Arrhythmias and mortality in conges- tive heart failure. The American journal of cardiology , 65(19):42–48, 1990. 5

    1990

  20. [28]

    Imagenet classification with deep convolutional neural net- works

    Alex Krizhevsky, Ilya Sutskever, and Geoffrey E Hinton. Imagenet classification with deep convolutional neural net- works. Communications of the ACM, 60(6):84–90, 2017. 5

    2017

  21. [29]

    A simple feature augmentation for do- main generalization

    Pan Li, Da Li, Wei Li, Shaogang Gong, Yanwei Fu, and Tim- othy M Hospedales. A simple feature augmentation for do- main generalization. In Proceedings of the IEEE/CVF Inter- national Conference on Computer Vision, pages 8886–8895,

  22. [30]

    Bayesian invariant risk minimization

    Yong Lin, Hanze Dong, Hao Wang, and Tong Zhang. Bayesian invariant risk minimization. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 16021–16030, 2022. 8

    2022

  23. [31]

    Federated semi-supervised medical image classifica- tion via inter-client relation matching

    Quande Liu, Hongzheng Yang, Qi Dou, and Pheng-Ann Heng. Federated semi-supervised medical image classifica- tion via inter-client relation matching. In Medical Image Computing and Computer Assisted Intervention–MICCAI 2021: 24th International Conference, Strasbourg, France, Se...

    2021

  24. [32]

    Un- biased atlas formation via large deformations metric map- ping

    Peter Lorenzen, Brad C Davis, and Sarang Joshi. Un- biased atlas formation via large deformations metric map- ping. In Medical Image Computing and Computer-Assisted Intervention–MICCAI 2005: 8th International Conference, Palm Springs, CA, USA, October 26-29, 2005, Proceedings,...

    2005

  25. [33]

    Invariant causal representation learning for out-of-distribution generalization

    Chaochao Lu, Yuhuai Wu, Jos ´e Miguel Hern ´andez-Lobato, and Bernhard Sch ¨olkopf. Invariant causal representation learning for out-of-distribution generalization. In Interna- tional Conference on Learning Representations, 2021. 1, 8

    2021

  26. [34]

    Geodesic shooting for computational anatomy

    Michael I Miller, Alain Trouv ´e, and Laurent Younes. Geodesic shooting for computational anatomy. Journal of mathematical imaging and vision, 24(2):209–228, 2006. 2

    2006

  27. [35]

    Representation learn- ing via invariant causal mechanisms

    Jovana Mitrovic, Brian McWilliams, Jacob Walker, Lars Buesing, and Charles Blundell. Representation learn- ing via invariant causal mechanisms. arXiv preprint arXiv:2010.07922, 2020. 1

    2010 arXiv

  28. [36]

    Learning from failure: De-biasing classifier from biased classifier

    Junhyun Nam, Hyuntak Cha, Sungsoo Ahn, Jaeho Lee, and Jinwoo Shin. Learning from failure: De-biasing classifier from biased classifier. Advances in Neural Information Pro- cessing Systems, 33:20673–20684, 2020. 1

    2020

  29. [37]

    Numerical optimiza- tion

    Jorge Nocedal and Stephen J Wright. Numerical optimiza- tion. Springer, 1999. 4

    1999

  30. [38]

    Segmentation, registration, and measurement of shape variation via im- age object shape

    Stephen M Pizer, Daniel S Fritsch, Paul A Yushkevich, Valen E Johnson, and Edward L Chaney. Segmentation, registration, and measurement of shape variation via im- age object shape. IEEE transactions on medical imaging , 18(10):851–865, 1999. 2

    1999

  31. [39]

    Ventricular arrhythmia in congestive heart failure

    Philip J Podrid, Richard I Fogel, and Therese Tordjman Fuchs. Ventricular arrhythmia in congestive heart failure. The American journal of cardiology, 69(18):82–96, 1992. 5

    1992

  32. [40]

    Age, alzheimer disease, and brain structure.Neu- rology, 73(22):1899–1905, 2009

    Cyrius A Raji, OL Lopez, LH Kuller, OT Carmichael, and JT Becker. Age, alzheimer disease, and brain structure.Neu- rology, 73(22):1899–1905, 2009. 5

    1905

  33. [41]

    Deep structural causal shape models

    Rajat Rasal, Daniel C Castro, Nick Pawlowski, and Ben Glocker. Deep structural causal shape models. In European Conference on Computer Vision , pages 400–432. Springer,

  34. [42]

    Deformable Image Registration for Surgical Guidance using Intraoperative Cone-Beam CT

    Sureerat Reaungamornrat. Deformable Image Registration for Surgical Guidance using Intraoperative Cone-Beam CT. PhD thesis, Johns Hopkins University, 2017. 1

    2017

  35. [43]

    Within-subject template estimation for unbi- ased longitudinal image analysis

    Martin Reuter, Nicholas J Schmansky, H Diana Rosas, and Bruce Fischl. Within-subject template estimation for unbi- ased longitudinal image analysis. Neuroimage, 61(4):1402– 1418, 2012. 5

    2012

  36. [44]

    U- net: Convolutional networks for biomedical image segmen- tation

    Olaf Ronneberger, Philipp Fischer, and Thomas Brox. U- net: Convolutional networks for biomedical image segmen- tation. In Medical Image Computing and Computer-Assisted Intervention–MICCAI 2015: 18th International Conference, Munich, Germany, October 5-9, 2015, Proceedings, Pa...

    2015

  37. [45]

    Very deep convo- lutional networks for large-scale image recognition

    Karen Simonyan and Andrew Zisserman. Very deep convo- lutional networks for large-scale image recognition. arXiv preprint arXiv:1409.1556, 2014. 3, 5

    2014 arXiv

  38. [46]

    Return of frus- tratingly easy domain adaptation

    Baochen Sun, Jiashi Feng, and Kate Saenko. Return of frus- tratingly easy domain adaptation. InProceedings of the AAAI conference on artificial intelligence, volume 30, 2016. 2

    2016

  39. [47]

    Deep coral: Correlation alignment for deep domain adaptation

    Baochen Sun and Kate Saenko. Deep coral: Correlation alignment for deep domain adaptation. In Computer Vision– ECCV 2016 Workshops: Amsterdam, The Netherlands, Oc- tober 8-10 and 15-16, 2016, Proceedings, Part III 14, pages 443–450. Springer, 2016. 2, 5, 7

    2016

  40. [48]

    Domain adaptation with conditional dis- tribution matching and generalized label shift

    Remi Tachet des Combes, Han Zhao, Yu-Xiang Wang, and Geoffrey J Gordon. Domain adaptation with conditional dis- tribution matching and generalized label shift. Advances in Neural Information Processing Systems , 33:19276–19289,

  41. [49]

    Causality-aware con- volutional neural networks for advanced image classification and generation

    Vagan Terziyan and Oleksandra Vitko. Causality-aware con- volutional neural networks for advanced image classification and generation. Procedia Computer Science, 217:495–506,

  42. [50]

    Statistics on diffeomorphisms via tangent space rep- resentations

    Marc Vaillant, Michael I Miller, Laurent Younes, and Alain Trouv´e. Statistics on diffeomorphisms via tangent space rep- resentations. NeuroImage, 23:S161–S169, 2004. 2

    2004

  43. [51]

    Principles of risk minimization for learn- ing theory

    Vladimir Vapnik. Principles of risk minimization for learn- ing theory. Advances in neural information processing sys- tems, 4, 1991. 4, 5

    1991

  44. [52]

    Diffeomorphic 3d image registration via geodesic shooting using an efficient adjoint calculation

    Franc ¸ois-Xavier Vialard, Laurent Risser, Daniel Rueckert, and Colin J Cotter. Diffeomorphic 3d image registration via geodesic shooting using an efficient adjoint calculation. International Journal of Computer Vision , 97(2):229–241,

  45. [53]

    Detect and correct bias in multi-site neuroimaging datasets.Medical Image Analysis, 67:101879, 2021

    Christian Wachinger, Anna Rieckmann, Sebastian P ¨olsterl, Alzheimer’s Disease Neuroimaging Initiative, et al. Detect and correct bias in multi-site neuroimaging datasets.Medical Image Analysis, 67:101879, 2021. 8

    2021

  46. [54]

    On calibration and out-of-domain generalization

    Yoav Wald, Amir Feder, Daniel Greenfeld, and Uri Shalit. On calibration and out-of-domain generalization. Advances in neural information processing systems , 34:2215–2227,

  47. [55]

    Geo-sic: Learning de- formable geometric shapes in deep image classifiers

    Jian Wang and Miaomiao Zhang. Geo-sic: Learning de- formable geometric shapes in deep image classifiers. The Conference on Neural Information Processing Systems ,

  48. [56]

    Out-of-distribution generalization with causal invari- ant transformations

    Ruoyu Wang, Mingyang Yi, Zhitang Chen, and Shengyu Zhu. Out-of-distribution generalization with causal invari- ant transformations. In Proceedings of the IEEE/CVF Con- ference on Computer Vision and Pattern Recognition, pages 375–385, 2022. 1, 8

    2022

  49. [57]

    Ai based cmr assessment of biventricular function: clinical sig- nificance of intervendor variability and measurement errors

    Shuo Wang, Hena Patel, Tamari Miller, Keith Ameyaw, Akhil Narang, Daksh Chauhan, Simran Anand, Emeka Anyanwu, Stephanie A Besser, Keigo Kawaji, et al. Ai based cmr assessment of biventricular function: clinical sig- nificance of intervendor variability and measurement errors. ...

    2022

  50. [58]

    Convolutional neural networks for classifica- tion of alzheimer’s disease: Overview and reproducible eval- uation

    Junhao Wen, Elina Thibeau-Sutre, Mauricio Diaz-Melo, Jorge Samper-Gonz´alez, Alexandre Routier, Simona Bottani, Didier Dormont, Stanley Durrleman, Ninon Burgos, Olivier Colliot, et al. Convolutional neural networks for classifica- tion of alzheimer’s disease: Overview and repr...

    2020

  51. [59]

    Treatment learning causal transformer for noisy image classification

    Chao-Han Huck Yang, I-Te Hung, Yi-Chieh Liu, and Pin- Yu Chen. Treatment learning causal transformer for noisy image classification. In Proceedings of the IEEE/CVF Win- ter Conference on Applications of Computer Vision , pages 6139–6150, 2023. 1, 2

    2023

  52. [60]

    Adversarial teacher-student representa- tion learning for domain generalization

    Fu-En Yang, Yuan-Chia Cheng, Zu-Yun Shiau, and Yu- Chiang Frank Wang. Adversarial teacher-student representa- tion learning for domain generalization. Advances in Neural Information Processing Systems, 34:19448–19460, 2021. 2

    2021

  53. [61]

    Improving out-of-distribution robustness via selective augmentation

    Huaxiu Yao, Yu Wang, Sai Li, Linjun Zhang, Weixin Liang, James Zou, and Chelsea Finn. Improving out-of-distribution robustness via selective augmentation. In International Con- ference on Machine Learning, pages 25407–25437. PMLR,

  54. [62]

    Atlas pre-selection strategies to enhance the efficiency and accuracy of multi-atlas brain seg- mentation tools

    Chenfei Ye, Ting Ma, Dan Wu, Can Ceritoglu, Michael I Miller, and Susumu Mori. Atlas pre-selection strategies to enhance the efficiency and accuracy of multi-atlas brain seg- mentation tools. PloS one, 13(7):e0200294, 2018. 3

    2018

  55. [63]

    Interventional few-shot learning

    Zhongqi Yue, Hanwang Zhang, Qianru Sun, and Xian-Sheng Hua. Interventional few-shot learning. Advances in neural information processing systems, 33:2734–2746, 2020. 1

    2020

  56. [64]

    Removal of con- founders via invariant risk minimization for medical diag- nosis

    Samira Zare and Hien Van Nguyen. Removal of con- founders via invariant risk minimization for medical diag- nosis. In Medical Image Computing and Computer Assisted Intervention–MICCAI 2022: 25th International Conference, Singapore, September 18–22, 2022, Proceedings, Part VIII...

    2022

  57. [65]

    Visualizing and un- derstanding convolutional networks

    Matthew D Zeiler and Rob Fergus. Visualizing and un- derstanding convolutional networks. In Computer Vision– ECCV 2014: 13th European Conference, Zurich, Switzer- land, September 6-12, 2014, Proceedings, Part I 13 , pages 818–833. Springer, 2014. 7

    2014

  58. [66]

    Causal intervention for weakly- supervised semantic segmentation

    Dong Zhang, Hanwang Zhang, Jinhui Tang, Xian-Sheng Hua, and Qianru Sun. Causal intervention for weakly- supervised semantic segmentation. Advances in Neural In- formation Processing Systems, 33:655–666, 2020. 1

    2020

  59. [67]

    Statistical shape anal- ysis: From landmarks to diffeomorphisms, 2016

    Miaomiao Zhang and Polina Golland. Statistical shape anal- ysis: From landmarks to diffeomorphisms, 2016. 1

    2016

  60. [68]

    Training confounder-free deep learning models for medical applica- tions

    Qingyu Zhao, Ehsan Adeli, and Kilian M Pohl. Training confounder-free deep learning models for medical applica- tions. Nature communications, 11(1):6010, 2020. 1, 2, 5, 7

    2020

  61. [69]

    Unet++: A nested u-net architecture for medical image segmentation

    Zongwei Zhou, Md Mahfuzur Rahman Siddiquee, Nima Tajbakhsh, and Jianming Liang. Unet++: A nested u-net architecture for medical image segmentation. In Deep learn- ing in medical image analysis and multimodal learning for clinical decision support, pages 3–11. Springer, 2018. 3

    2018

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.