Pith. sign in

REVIEW 4 major objections 6 minor 51 references

DUAL-GLOW: Conditional Flow-Based Generative Model for Modality Transfer

T0 review · 4 major / 6 minor · reviewed 2026-08-14 · deepseek-v4-flash

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

Pith's one-line read DUAL-GLOW is a flow-based generative model that learns the conditional distribution of PET given MRI and reports synthetic 3D volumes that outperform GAN-based baselines on the ADNI dataset.

desk verdict The architectural idea and derivation are solid, but the empirical headline overreaches: the numbers in Table 1 don't support 'quantitatively better' without significance tests. read the letter →

arxiv 1908.08074 v1 pith:ZHARS65T submitted 2019-08-21 eess.IV cs.CV

Haoliang Sun , Ronak Mehta , Hao H. Zhou , Zhichun Huang , Sterling C. Johnson , Vivek Prabhakaran , Vikas Singh This is my paper · ORCID
classification eess.IVcs.CV
keywords flow-basedgenerativemodelmodalitytransferMRI-to-PETsynthesisnormalizingflowsconditionallikelihoodAlzheimer'sdiseaseFDG-PETsideinformationconditioning
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

The paper tries to establish that flow-based generative models can perform MRI-to-PET modality transfer in a small-sample medical setting by learning the conditional distribution of PET given MRI in latent space. It argues that with two invertible networks and a relation network, maximizing conditional likelihood reduces to a tractable objective and yields sharper, more faithful synthetic PET than adversarial and autoencoder baselines. It also claims that side information such as age can be disentangled into the top latent level, so the model can generate PET images that reflect age-related hypometabolism. If these claims hold, PET synthesis becomes a likelihood-based, invertible procedure with exact latent inference and useful downstream diagnostic signal.

What carries the argument

The load-bearing object is a paired-flow likelihood: two invertible affine-coupling flows with multi-scale splitting, and a relation network that parameterizes the Gaussian conditional density p(z_p|z_m) = N(z_p; mu_theta(z_m), sigma_theta(z_m)). The identity log p(x_p|x_m) = log p(z_p|z_m) + log|det(dz_p/dx_p)| lets the model train with exact log-likelihood; affine coupling makes the Jacobian log-determinant a simple sum over scale terms, and multi-scale splitting reduces computation. For side information, gradient-reversal-layer-equipped discriminators strip age or attribute signal from lower latent levels while a top-level discriminator preserves it, enforcing the assumption that only the highest level of the representation is affected by the conditioning variable.

What would settle it

Train an age predictor on the lower-level MRI latent codes of the age-conditioned model; if its accuracy is well above chance, the top-level-only assumption is violated, and similarly, replacing the Gaussian conditional with a more flexible density and showing materially better held-out conditional log-likelihood would falsify the Gaussianity assumption.

Watch

Extended reading notes

Core claim

DUAL-GLOW's central claim is that the conditional distribution of PET given MRI can be learned with two normalizing flows, one mapping PET to a latent code and one mapping MRI to its own latent code, plus a relation network that predicts a Gaussian conditional density between the two latent spaces. Under the change-of-variables rule, the conditional log-likelihood of PET given MRI equals the log of this latent conditional density plus the log-determinant of the PET flow's Jacobian, and adding a regularizer on the MRI marginal gives the training objective in Eq. (12). The paper reports that on 806 ADNI MRI/PET pairs, DUAL-GLOW produces full 3D PET volumes with higher SSIM and PSNR and lower MAE than cGAN, UcGAN, C-VAE and pix2pix, and that an SVM trained on synthetic PET achieves AD/CN classification accuracy comparable to ground truth. The age-conditioned extension shows decreasing regional intensity with age, matching the expected aging-related hypometabolism.

Load-bearing premise

The whole derivation assumes that, after the two flows, the conditional density of the PET latent given the MRI latent is exactly Gaussian with mean and variance produced by neural networks, and the age-conditioning version further assumes side information lives only in the top latent level; the paper gives no direct evidence for either assumption.

Editorial extensions

If this is right

  • If DUAL-GLOW is correct, MRI-to-PET synthesis can be trained by maximizing a tractable conditional likelihood rather than by adversarial or reconstruction losses, which avoids the blur and mode collapse associated with those approaches.
  • Because both transformations are invertible, the same trained pair of flows provides exact latent-variable inference and lets a practitioner sample diverse PET volumes for a single MRI rather than one deterministic output.
  • The age-conditioned extension implies that, for a fixed MRI, PET-like hypometabolism changes with the side variable can be simulated, giving a tool for studying progression of neurodegeneration across the age range.
  • Downstream diagnostic classifiers trained on synthetic PET retain most of the discriminative signal of real PET, with the paper reporting 91% versus 94% accuracy for AD/CN classification, so generated volumes could augment small cohorts in statistical analyses.

Reading between the lines

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

  • Because the two flows are invertible, the framework could in principle run in reverse to estimate plausible MRI volumes from PET, a direction the paper does not test.
  • The Gaussian conditional assumption is not directly validated; an immediate extension would be replacing p(z_p|z_m) with a conditional normalizing flow on the latent pair and measuring whether held-out conditional log-likelihood or synthesis quality improves.
  • The GRL-disentanglement design could be applied to other side variables such as disease status, sex, or genotype, though the paper's wide variance bands suggest larger sample sizes are needed for reliable group comparisons.
  • The relation network's Gaussian parameters provide a natural source of uncertainty: repeated sampling from p(z_p|z_m) could yield voxel-level confidence intervals for metabolism, which the paper does not report.
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 / 6 minor

Summary. The paper introduces DUAL-GLOW, a normalizing-flow model for conditional MRI-to-PET synthesis. Two invertible networks map MRI and PET into latent spaces, and a relation network models the conditional density of the PET latent given the MRI latent as a Gaussian whose mean and variance are network outputs. The conditional log-likelihood is derived via the change-of-variables formula in Eqs. (8)-(12). The framework is extended to side information such as age by adding discriminators and gradient reversal layers intended to confine the effect of the covariate to the top-most latent level. Experiments on ADNI compare DUAL-GLOW with cGAN, UcGAN, C-VAE, and pix2pix using MAE, CorCoef, PSNR, and SSIM, and include an SVM classification of generated images and age-conditioned PET generation.

Significance. If the claims hold, the paper provides a useful likelihood-based alternative to GANs for medical modality transfer, with the practical advantage of exact latent inference and full 3D-volume processing. The mathematical derivation in Section 3 is clean and self-contained: Eq. (10) correctly follows from the block-diagonal Jacobian in Eq. (11), and Eq. (12) is a valid conditional-likelihood objective with a regularizer on the MRI marginal. The release of code is a concrete strength. However, the empirical support for the headline superiority claim is incomplete: no significance tests are reported, the cross-validation protocol is ambiguous, one of the four headline metrics (CorCoef) is below the C-VAE baseline, and the age-conditioning results are qualitative with wide variance bands. The significance of the paper is therefore moderate pending strengthened statistical evidence and a more measured abstract.

major comments (4)
  1. [Section 4.2, Table 1] The abstract's claim that DUAL-GLOW is 'quantitatively better than recent works' is not supported by the reported statistics. DUAL-GLOW has CorCoef 0.975 versus C-VAE 0.980, so it is numerically worse on one of the four headline metrics. For PSNR, the means are 29.56±2.66 versus 28.69±2.06, and for SSIM 0.898±0.06 versus 0.817±0.06; the one-standard-deviation intervals for PSNR overlap substantially, and no confidence intervals, paired tests, or per-fold results are provided. Please add a properly paired significance analysis across the folds or subjects with multiple-comparison control and report effect sizes, or revise the abstract to 'comparable or better' with appropriate caveats.
  2. [Section 4.1] The sentence 'randomly select 726 subjects as the training data and the remaining 80 as testing within a 10-fold evaluation scheme' is ambiguous. A standard 10-fold cross-validation on 806 subjects would use test sets of roughly 80 subjects, but a single random 80-subject split is not 10-fold. Please specify exactly how many folds or splits were used, whether the same folds were used for all compared methods, and how test subjects were selected for the age-conditioning experiments; this determines the effective number of independent test evaluations and the validity of any significance statements.
  3. [Section 3, 'How to condition based on side information'; Figure 7] The age-conditioning contribution rests on the assumption that side information affects only the top-most latent level and that gradient reversal removes age from the lower levels. No quantitative evidence is given that the lower-level latents are indeed age-invariant, for example via classifier accuracy on those latents before and after the gradient reversal layers. The only quantitative support is Figure 7, whose 95% bands are described by the authors as too wide for statistically sound conclusions. The abstract's statement that the model can 'capture brain FDG-PET changes as a function of age' should therefore be backed by a statistical test of the age trend with confidence intervals and a validation of the disentanglement assumption, or softened to a qualitative claim.
  4. [Section 3, Eq. (6)] The objective in Eq. (12) is exactly the conditional log-likelihood only under the Gaussian family pθ(zp|zm)=N(zp;µθ(zm),σθ(zm)). The paper provides no diagnostic for this assumption. Please include a residual analysis on held-out subjects, for example standardized residuals (zp−µθ(zm))/σθ(zm) compared with a standard normal, or a comparison against a more flexible conditional density, or explicitly discuss the limitation. This is a correctness-risk check on the central likelihood interpretation rather than a request for a different model.
minor comments (6)
  1. [Abstract and Section 4.1] The dataset size is given as 826 subjects in the abstract but 806 clean MRI/PET pairs in Section 4.1; please reconcile these numbers.
  2. [Section 4.1] The optimizer is called 'AdamMax' but should be 'AdaMax' to match the terminology in reference [20].
  3. [Figure 3 caption] The caption contains a typo: 'Spliting' should be 'Splitting'.
  4. [Section 4.2] The metric name 'Cor Coef' should be written consistently as 'CorCoef' as in Table 1.
  5. [Figure 7] The x-axis is labeled only '50 100'; please label the axis 'Age' with explicit tick values and add units.
  6. [Equation (15)] The notation I1:d1 is not defined; it should denote the identity matrix acting on the first d1 components.

Circularity Check

1 steps flagged · score 2.0 of 10

Core flow-based conditional derivation is self-contained; only the age-conditioning demonstration restates its own conditioning input.

  1. fitted input called prediction [Sec. 3, Eq. (16)-(17); Sec. 4.3 'Adjusting for Age with Conditioning'; Fig. 7]
    "Training the full conditional DUAL-GLOW model, we use ground truth 'side' information (age) as the conditioning variable described above. ... As we see in Figure 7, the mean intensity values show a downward trend with age, as expected."

    Age is an explicit conditioning input c in Eq. (16)-(17): the objective maximizes log pθ(zp|z′_m,c) with pθ(zp|z′_m,c)=N(zp; µθ(z′_m,c), σθ(z′_m,c)). The paper trains the full conditional DUAL-GLOW model on ground-truth age as the conditioning variable, then reports that generated ROI intensities show a downward trend with age. That trend is a property of the fitted conditional generator, not an independent derivation; the abstract's claim that the model 'capture[s] brain FDG-PET changes ... as a function of age' restates the conditioning setup rather than predicting an unseen relationship. This affects only the age-conditional demonstration; the base MRI-to-PET likelihood derivation in Eqs. (8)-(12) is self-contained.

full rationale

The central derivation is self-contained. With invertible flows fm and fp, Eq. (8)-(12) correctly factorize the conditional log-likelihood into a latent conditional log p(zp|zm), the PET-flow log-Jacobian, and a regularized MR marginal; none of these quantities is defined in terms of the target PET output beyond the ordinary likelihood construction. The relation network is a learned conditional Gaussian in latent space, an explicit modeling assumption, not a recycled result. There is no load-bearing self-citation: the cited flows (RealNVP, GLOW) and GRL are external prior work, and the authors' own listed prior papers are unrelated to modality transfer. The main empirical claim is benchmarked externally against cGAN, UcGAN, C-VAE, and pix2pix on ADNI, so it is not forced by a fitted parameter; the unsupported 'quantitatively better' wording in the abstract is a statistical/correctness concern, not a circularity one. The one mild circularity is the age-conditioning demonstration: age is fed in as the conditioning variable c in Eq. (16)-(17), and the resulting age-related decline is presented as the model capturing hypometabolism with age, which is a sanity check of the learned regression rather than an independent finding. The paper itself notes that the wide variance bands mean 'a larger sample size may be necessary to derive statistically sound conclusions,' underscoring the weakness of that particular claim. Overall, the score is low because the core synthesis result has independent content and external benchmarks.

Assumptions & free parameters 3 free parameters · 6 assumptions · 0 invented entities

The model pulls heavily from GLOW and RealNVP for the flow layers and from GRL for the disentanglement, and it assumes Gaussian latent conditionals. The main parameters chosen by hand are the regularizer lambda, the classifier loss weight, and the network architecture. The side-information disentanglement is the most ad hoc component.

free parameters (3)
  • lambda (MRI marginal regularizer weight) = 0.001
    Set by hand in all experiments; Appendix 6.4 shows lambda=1.0 was tried and produced worse qualitative results.
  • classification loss weight = 0.01
    Set by hand for all discriminator losses in the conditional framework; no ablation is reported.
  • network depth and width = 4 levels, 16 coupling layers per level, 512 intermediate channels
    Architecture chosen by hand for the ADNI experiments; no systematic study of these choices is reported.
assumptions (6)
  • standard math Change-of-variables formula for invertible transformations holds for the flow maps.
    Used in Eq. (4) and Eq. (10) to compute log-likelihoods.
  • standard math Affine coupling layers are invertible and have tractable log-determinants.
    Inherited from RealNVP and GLOW, applied in Section 3 for the flow functions f_p and f_m.
  • domain assumption Paired MRI and PET images are accurately coregistered and normalized after SPM12 preprocessing.
    Section 4.1 describes alignment to MNI152 space; registration errors would directly corrupt the conditional mapping.
  • domain assumption The conditional distribution of the PET latent given the MRI latent is Gaussian with neural-network mean and variance.
    Stated in Eq. (6) and used throughout the derivation; no evidence is provided that this family is rich enough.
  • ad hoc to paper Side information such as age affects only the top-most latent level and can be removed from lower levels using gradient reversal.
    Stated as an expectation in Section 3: 'we expect that only the highest level of DUAL-GLOW should be affected by this'.
  • domain assumption The hierarchical splitting technique preserves enough information for accurate generation.
    Figure 3 and Section 3 adopt the multi-scale splitting from prior flow literature; the paper does not ablate this choice.

how reviews work

0 comments
Cite this review

Pith. "Pith review of DUAL-GLOW: Conditional Flow-Based Generative Model for Modality Transfer." pith.science (2026). https://pith.science/paper/ZHARS65T

@misc{pith2026190808074,
  author       = {Pith},
  title        = {Pith review of: DUAL-GLOW: Conditional Flow-Based Generative Model for Modality Transfer},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ZHARS65T}},
  note         = {Machine review of arXiv:1908.08074}
}
read the original abstract

Positron emission tomography (PET) imaging is an imaging modality for diagnosing a number of neurological diseases. In contrast to Magnetic Resonance Imaging (MRI), PET is costly and involves injecting a radioactive substance into the patient. Motivated by developments in modality transfer in vision, we study the generation of certain types of PET images from MRI data. We derive new flow-based generative models which we show perform well in this small sample size regime (much smaller than dataset sizes available in standard vision tasks). Our formulation, DUAL-GLOW, is based on two invertible networks and a relation network that maps the latent spaces to each other. We discuss how given the prior distribution, learning the conditional distribution of PET given the MRI image reduces to obtaining the conditional distribution between the two latent codes w.r.t. the two image types. We also extend our framework to leverage 'side' information (or attributes) when available. By controlling the PET generation through 'conditioning' on age, our model is also able to capture brain FDG-PET (hypometabolism) changes, as a function of age. We present experiments on the Alzheimers Disease Neuroimaging Initiative (ADNI) dataset with 826 subjects, and obtain good performance in PET image synthesis, qualitatively and quantitatively better than recent works.

Figures

Figures reproduced from arXiv: 1908.08074 by the authors.

Figure 1
Figure 1. The DUAL-GLOW framework. For the conditional module, the dashed and dotted pieces are added and removed [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. DUAL-GLOW for image generation [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 4
Figure 4. Box plot of MAE metrics for different methods. [PITH_FULL_IMAGE:figures/full_fig_p006_4.png] view at source ↗
Figures from the paper (13 more)
Figure 5
Figure 5. Figure 5: Synthetic images are meaningful for subjects in both extremes of disease spectrum. Left: CN. Middle: MCI. Right: AD. The generated PET images show consistency of hypometabolism (less red, more yellow) with the ground truth image. (Best viewed in color; montages shown i…
Figure 6
Figure 6. Figure 6: Conditioning on age should yield generated images that show increased hypometabolism with age. These are representative results from our PET generation as a function of age. As we scan left to right, we indeed see a decrease in metabolism (less red, more yellow) which …
Figure 7
Figure 7. Figure 7: The mean intensity with 95% standard devia [PITH_FULL_IMAGE:figures/full_fig_p008_7.png]
Figure 8
Figure 8. Figure 8: Sample generation using DUAL-GLOW. The first [PITH_FULL_IMAGE:figures/full_fig_p008_8.png]
Figure 17
Figure 17. Figure 17: Better for the small λ. 12 [PITH_FULL_IMAGE:figures/full_fig_p012_17.png]
Figure 9
Figure 9. Figure 9: Two synthetic CN subjects. From left to right: input of MRI, the ground truth, the synthetic subject. [PITH_FULL_IMAGE:figures/full_fig_p013_9.png]
Figure 10
Figure 10. Figure 10: Two synthetic MCI subjects. From left to right: input of MRI, the ground truth, the synthetic subject. [PITH_FULL_IMAGE:figures/full_fig_p013_10.png]
Figure 11
Figure 11. Figure 11: Two synthetic AD subjects. From left to right: input of MRI, the ground truth, the synthetic subject. [PITH_FULL_IMAGE:figures/full_fig_p014_11.png]
Figure 12
Figure 12. Figure 12: DUAL-GLOW can produce more accurate prediction in dashed rectangles. [PITH_FULL_IMAGE:figures/full_fig_p014_12.png]
Figure 13
Figure 13. Figure 13: Age information manipulation. There are 3 subjects, AD, MCI, and CN, each subject provides [PITH_FULL_IMAGE:figures/full_fig_p015_13.png]
Figure 14
Figure 14. Figure 14: Decreasing trends for 30 ROIs (related to aging). 16 [PITH_FULL_IMAGE:figures/full_fig_p016_14.png]
Figure 15
Figure 15. Figure 15: Input cartoon images and generated/reconstructed faces applying our DUAL-GLOW framework to the CelebA [PITH_FULL_IMAGE:figures/full_fig_p017_15.png]
Figure 16
Figure 16. Figure 16: Input “sketch” images and generated/reconstructed shoes applying our DUAL-GLOW framework to the UT [PITH_FULL_IMAGE:figures/full_fig_p017_16.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

51 extracted references · 34 canonical work pages

  1. [1]

    Wasserstein gan

    Martin Arjovsky, Soumith Chintala, and L ´eon Bottou. Wasserstein gan. arXiv preprint arXiv:1701.07875, 2017

    arXiv 2017

  2. [2]

    Spm12 man- ual

    John Ashburner, Gareth Barnes, C Chen, Jean Daunizeau, Guillaume Flandin, Karl Friston, Stefan Kiebel, James Kil- ner, Vladimir Litvak, Rosalyn Moran, et al. Spm12 man- ual. Wellcome Trust Centre for Neuroimaging, London, UK, 2014

    work page 2014

  3. [3]

    Large scale gan training for high fidelity natural image synthesis

    Andrew Brock, Jeff Donahue, and Karen Simonyan. Large scale gan training for high fidelity natural image synthesis. arXiv preprint arXiv:1809.11096, 2018

    arXiv 2018

  4. [4]

    Mode regularized generative adversarial net- works

    Tong Che, Yanran Li, Athul Paul Jacob, Yoshua Bengio, and Wenjie Li. Mode regularized generative adversarial net- works. arXiv preprint arXiv:1612.02136, 2016

    arXiv 2016

  5. [5]

    Training deep nets with sublinear memory cost

    Tianqi Chen, Bing Xu, Chiyuan Zhang, and Carlos Guestrin. Training deep nets with sublinear memory cost. arXiv preprint arXiv:1604.06174, 2016

    arXiv 2016

  6. [6]

    Stargan: Unified genera- tive adversarial networks for multi-domain image-to-image translation

    Yunjey Choi, Minje Choi, Munyoung Kim, Jung-Woo Ha, Sunghun Kim, and Jaegul Choo. Stargan: Unified genera- tive adversarial networks for multi-domain image-to-image translation. In Proceedings of the IEEE Conference on Com- puter Vision and Pattern Recognition , pages 8789–8797, 2018

    work page 2018

  7. [7]

    Nice: Non-linear independent components estimation

    Laurent Dinh, David Krueger, and Yoshua Bengio. Nice: Non-linear independent components estimation. arXiv preprint arXiv:1410.8516, 2014

    arXiv 2014

  8. [8]

    Density estimation using real nvp

    Laurent Dinh, Jascha Sohl-Dickstein, and Samy Ben- gio. Density estimation using real nvp. arXiv preprint arXiv:1605.08803, 2016

    arXiv 2016

Show all 51 references
  1. [9]

    A varia- tional u-net for conditional appearance and shape generation

    Patrick Esser, Ekaterina Sutter, and Bj ¨orn Ommer. A varia- tional u-net for conditional appearance and shape generation. In Computer Vision and Pattern Recognition , pages 8857– 8866, 2018

    2018

  2. [10]

    Unsupervised domain adaptation by backpropagation

    Yaroslav Ganin and Victor Lempitsky. Unsupervised domain adaptation by backpropagation. arXiv preprint arXiv:1409.7495, 2014

    2014 arXiv

  3. [11]

    Image-to-image translation for cross-domain disentan- glement

    Abel Gonzalez-Garcia, Joost van de Weijer, and Yoshua Ben- gio. Image-to-image translation for cross-domain disentan- glement. In Advances in Neural Information Processing Sys- tems, pages 1294–1305, 2018

    2018

  4. [12]

    Generative adversarial nets

    Ian Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron Courville, and Yoshua Bengio. Generative adversarial nets. In Advances in neural information processing systems, pages 2672–2680, 2014

    2014

  5. [13]

    Mr-based synthetic ct generation using a deep convolutional neural network method

    Xiao Han. Mr-based synthetic ct generation using a deep convolutional neural network method. Medical physics , 44(4):1408–1419, 2017

    2017

  6. [14]

    Mkl for robust multi-modality ad classification

    Chris Hinrichs, Vikas Singh, Guofan Xu, and Sterling John- son. Mkl for robust multi-modality ad classification. In In- ternational Conference on Medical Image Computing and Computer-Assisted Intervention , pages 786–794. Springer, 2009

    2009

  7. [15]

    Multimodal unsupervised image-to-image translation

    Xun Huang, Ming-Yu Liu, Serge Belongie, and Jan Kautz. Multimodal unsupervised image-to-image translation. In European Conference on Computer Vision, pages 172–189, 2018

    2018

  8. [16]

    Image-to-image translation with conditional adversar- ial networks

    Phillip Isola, Jun-Yan Zhu, Tinghui Zhou, and Alexei A Efros. Image-to-image translation with conditional adversar- ial networks. In Computer Vision and Pattern Recognition, pages 1125–1134, 2017

    2017

  9. [17]

    Prediction of standard-dose brain pet image by using mri and low-dose brain [18f] fdg pet im- ages

    Jiayin Kang, Yaozong Gao, Feng Shi, David S Lalush, Weili Lin, and Dinggang Shen. Prediction of standard-dose brain pet image by using mri and low-dose brain [18f] fdg pet im- ages. Medical physics, 42(9):5301–5309, 2015

    2015

  10. [18]

    Progressive growing of gans for improved quality, stability, and variation

    Tero Karras, Timo Aila, Samuli Laine, and Jaakko Lehtinen. Progressive growing of gans for improved quality, stability, and variation. arXiv preprint arXiv:1710.10196, 2017

    2017 arXiv

  11. [19]

    Learning to discover cross-domain relations with generative adversarial networks

    Taeksoo Kim, Moonsu Cha, Hyunsoo Kim, Jung Kwon Lee, and Jiwon Kim. Learning to discover cross-domain relations with generative adversarial networks. In International Con- ference on Machine Learning, pages 1857–1865. JMLR. org, 2017

    2017

  12. [20]

    Adam: A method for stochastic optimization

    Diederik P Kingma and Jimmy Ba. Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980 , 2014

    2014 arXiv

  13. [21]

    Glow: Generative flow with invertible 1x1 convolutions

    Durk P Kingma and Prafulla Dhariwal. Glow: Generative flow with invertible 1x1 convolutions. InAdvances in Neural Information Processing Systems, pages 10235–10244, 2018

    2018

  14. [22]

    Diverse image-to-image translation via disentangled representations

    Hsin-Ying Lee, Hung-Yu Tseng, Jia-Bin Huang, Maneesh Singh, and Ming-Hsuan Yang. Diverse image-to-image translation via disentangled representations. In European Conference on Computer Vision, pages 35–51, 2018

    2018

  15. [23]

    Deep learning based imaging data completion for improved brain disease diag- nosis

    Rongjian Li, Wenlu Zhang, Heung-Il Suk, Li Wang, Jiang Li, Dinggang Shen, and Shuiwang Ji. Deep learning based imaging data completion for improved brain disease diag- nosis. In International Conference on Medical Image Com- puting and Computer-Assisted Intervention, pages 305...

    2014

  16. [24]

    Conditional image-to-image translation

    Jianxin Lin, Yingce Xia, Tao Qin, Zhibo Chen, and Tie-Yan Liu. Conditional image-to-image translation. In Computer Vision and Pattern Recognition, pages 5524–5532, 2018

    2018

  17. [25]

    Unsupervised image-to-image translation networks

    Ming-Yu Liu, Thomas Breuel, and Jan Kautz. Unsupervised image-to-image translation networks. In Advances in Neural Information Processing Systems, pages 700–708, 2017

    2017

  18. [26]

    Deep learning face attributes in the wild

    Ziwei Liu, Ping Luo, Xiaogang Wang, and Xiaoou Tang. Deep learning face attributes in the wild. In International Conference on Computer Vision, pages 3730–3738, 2015

    2015

  19. [27]

    Exemplar guided unsupervised image-to- image translation with semantic consistency

    Liqian Ma, Xu Jia, Stamatios Georgoulis, Tinne Tuytelaars, and Luc Van Gool. Exemplar guided unsupervised image-to- image translation with semantic consistency. InInternational Conference on Learning Representations, 2019

    2019

  20. [28]

    Dose evaluation of fast synthetic-ct generation using a generative adversarial network for general pelvis mr-only ra- diotherapy

    Matteo Maspero, Mark HF Savenije, Anna M Dinkla, Pe- ter R Seevinck, Martijn PW Intven, Ina M Jurgenliemk- Schulz, Linda GW Kerkmeijer, and Cornelis AT van den Berg. Dose evaluation of fast synthetic-ct generation using a generative adversarial network for general pelvis mr-on...

    2018

  21. [29]

    Conditional generative adversarial nets

    Mehdi Mirza and Simon Osindero. Conditional generative adversarial nets. arXiv preprint arXiv:1411.1784, 2014

    2014 arXiv

  22. [30]

    Instance-aware image-to-image translation

    Sangwoo Mo, Minsu Cho, and Jinwoo Shin. Instance-aware image-to-image translation. In International Conference on Learning Representations, 2019

    2019

  23. [31]

    Image to image transla- tion for domain adaptation

    Zak Murez, Soheil Kolouri, David Kriegman, Ravi Ra- mamoorthi, and Kyungnam Kim. Image to image transla- tion for domain adaptation. In Computer Vision and Pattern Recognition, pages 4500–4509, 2018

    2018

  24. [32]

    Medical image syn- thesis with context-aware generative adversarial networks

    Dong Nie, Roger Trullo, Jun Lian, Caroline Petitjean, Su Ruan, Qian Wang, and Dinggang Shen. Medical image syn- thesis with context-aware generative adversarial networks. In International Conference on Medical Image Computing and Computer-Assisted Intervention , pages 417–425....

    2017

  25. [33]

    Pixel recurrent neural networks

    Aaron van den Oord, Nal Kalchbrenner, and Koray Kavukcuoglu. Pixel recurrent neural networks. arXiv preprint arXiv:1601.06759, 2016

    2016 arXiv

  26. [34]

    Synthesizing missing pet from mri with cycle-consistent generative adversarial net- works for alzheimers disease diagnosis

    Yongsheng Pan, Mingxia Liu, Chunfeng Lian, Tao Zhou, Yong Xia, and Dinggang Shen. Synthesizing missing pet from mri with cycle-consistent generative adversarial net- works for alzheimers disease diagnosis. In International Conference on Medical Image Computing and Computer- As...

    2018

  27. [35]

    Varia- tional inference with normalizing flows

    Danilo Jimenez Rezende and Shakir Mohamed. Varia- tional inference with normalizing flows. arXiv preprint arXiv:1505.05770, 2015

    2015 arXiv

  28. [36]

    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 International Conference on Medical image com- puting and computer-assisted intervention , pages 234–241. Springer, 2015

    2015

  29. [37]

    Mri to fdg-pet: Cross-modal synthesis using 3d u-net for multi-modal alzheimers classification

    Apoorva Sikka, Skand Vishwanath Peri, and Deepti R Bathula. Mri to fdg-pet: Cross-modal synthesis using 3d u-net for multi-modal alzheimers classification. In Inter- national Workshop on Simulation and Synthesis in Medical Imaging, pages 80–89. Springer, 2018

    2018

  30. [38]

    Determinants of block matrices

    John R Silvester. Determinants of block matrices. The Math- ematical Gazette, 84(501):460–467, 2000

    2000

  31. [39]

    Learning structured output representation using deep conditional gen- erative models

    Kihyuk Sohn, Honglak Lee, and Xinchen Yan. Learning structured output representation using deep conditional gen- erative models. In Advances in neural information process- ing systems, pages 3483–3491, 2015

    2015

  32. [40]

    Direct estimation of spinal cobb angles by structured multi-output regression

    Haoliang Sun, Xiantong Zhen, Chris Bailey, Parham Ra- soulinejad, Yilong Yin, and Shuo Li. Direct estimation of spinal cobb angles by structured multi-output regression. In International Conference on Information Processing in Med- ical Imaging, pages 529–540. Springer, 2017

    2017

  33. [41]

    Learning deep match kernels for image-set classification

    Haoliang Sun, Xiantong Zhen, Yuanjie Zheng, Gongping Yang, Yilong Yin, and Shuo Li. Learning deep match kernels for image-set classification. In Computer Vision and Pattern Recognition, pages 3307–3316, 2017

    2017

  34. [42]

    Perceptual adversarial networks for image-to-image transformation

    Chaoyue Wang, Chang Xu, Chaohui Wang, and Dacheng Tao. Perceptual adversarial networks for image-to-image transformation. IEEE Transactions on Image Processing , 27(8):4066–4079, 2018

    2018

  35. [43]

    Mix and match networks: encoder-decoder alignment for zero- pair image translation

    Yaxing Wang, Joost van de Weijer, and Luis Herranz. Mix and match networks: encoder-decoder alignment for zero- pair image translation. In Proceedings of the IEEE Con- ference on Computer Vision and Pattern Recognition, pages 5467–5476, 2018

    2018

  36. [44]

    Deep mr to ct synthesis using unpaired data

    Jelmer M Wolterink, Anna M Dinkla, Mark HF Savenije, Pe- ter R Seevinck, Cornelis AT van den Berg, and Ivana Iˇsgum. Deep mr to ct synthesis using unpaired data. InInternational Workshop on Simulation and Synthesis in Medical Imaging , pages 14–23. Springer, 2017

    2017

  37. [45]

    Opti- mum template selection for atlas-based segmentation

    Minjie Wu, Caterina Rosano, Pilar Lopez-Garcia, Cameron S Carter, and Howard J Aizenstein. Opti- mum template selection for atlas-based segmentation. NeuroImage, 34(4):1612–1618, 2007

    2007

  38. [46]

    Holistically-nested edge de- tection

    Saining Xie and Zhuowen Tu. Holistically-nested edge de- tection. In International Conference on Computer Vision , pages 1395–1403, 2015

    2015

  39. [47]

    Fine-grained visual compar- isons with local learning

    Aron Yu and Kristen Grauman. Fine-grained visual compar- isons with local learning. In Computer Vision and Pattern Recognition, Jun 2014

    2014

  40. [48]

    Semantic jitter: Dense su- pervision for visual comparisons via synthetic images

    Aron Yu and Kristen Grauman. Semantic jitter: Dense su- pervision for visual comparisons via synthetic images. In International Conference on Computer Vision, Oct 2017

    2017

  41. [49]

    Unpaired image-to-image translation using cycle- consistent adversarial networks

    Jun-Yan Zhu, Taesung Park, Phillip Isola, and Alexei A Efros. Unpaired image-to-image translation using cycle- consistent adversarial networks. In International Conference on Computer Vision, pages 2223–2232, 2017

    2017

  42. [50]

    To- ward multimodal image-to-image translation

    Jun-Yan Zhu, Richard Zhang, Deepak Pathak, Trevor Dar- rell, Alexei A Efros, Oliver Wang, and Eli Shechtman. To- ward multimodal image-to-image translation. In Advances in Neural Information Processing Systems , pages 465–476, 2017. 10

    2017

  43. [51]

    lev- els

    Appendix In this supplement, We provide extensive additional de- tails regarding the neuroimaging experiments and some the- oretical analysis. 6.1. The ADNI dataset Data used in the experiments for this work were ob- tained directly from the Alzheimers Disease Neuroimaging Ini...

    2003

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.