Revisiting CycleGAN for semi-supervised segmentation

Aniket Agarwal; Arnab Kumar Mondal; Christian Desrosiers; Jose Dolz

REVIEW 4 major objections 5 minor 34 references

Revisiting CycleGAN for semi-supervised segmentation

T0 review · 4 major / 5 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 The paper claims that enforcing cycle consistency between unpaired images and segmentation masks adds an unsupervised regularization effect that improves segmentation accuracy when labeled data is scarce.

desk verdict A worthwhile empirical idea—cycle-consistent image↔mask translation as a regularizer for low-label segmentation—but the mechanism is asserted rather than demonstrated, and the headline 2-4% overstates the ACDC numbers. read the letter →

arxiv 1908.11569 v1 pith:UHPLKZRK submitted 2019-08-30 cs.CV

Arnab Kumar Mondal , Aniket Agarwal , Jose Dolz , Christian Desrosiers This is my paper · ORCID

classification cs.CV

keywords semanticsegmentationsemi-supervisedlearningCycleGANconsistencygenerativeadversarialnetworksunlabeleddatamedicalimaging

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 unpaired image-to-image translation, typically used for style transfer, can serve as a semi-supervised regularizer for semantic segmentation. Its model learns a two-way mapping between real images and segmentation masks, with cycle-consistency losses that let unlabeled images constrain the segmenter without needing annotations. Across PASCAL VOC 2012, Cityscapes, and ACDC, this raises mean intersection-over-union by 2-4 points over training on labeled images alone when only 10-20% of images are labeled. The gain shrinks to about 1 point at 50% labels. If true, this offers a practical way to exploit abundant unlabeled images in labeling-limited settings such as medical imaging and autonomous driving.

What carries the argument

The load-bearing mechanism is CycleGAN-style cycle consistency applied between the image domain and the segmentation-mask domain. The named components are two conditional generators, $G_{IS}$ from image to mask and $G_{SI}$ from mask to image, plus two patch-wise discriminators. The regularizing effect comes from the two cycle losses: an unlabeled image passed through $G_{IS}$ then $G_{SI}$ must return to itself under an L1 pixel loss, and a ground-truth mask passed through $G_{SI}$ then $G_{IS}$ must return to itself under cross-entropy. These cycles tie the segmenter's internal representation to invertible image structure, which is the paper's proposed unsupervised prior.

What would settle it

Train the same model on an unlabeled set whose images are patch-shuffled so that local texture and color are preserved but global object layout is destroyed. If the cycle-consistency gain over the partial-supervision baseline persists on this corrupted unlabeled set, the image-cycle loss is being satisfied through low-level statistics rather than semantics; if the gain vanishes, the regularizer depends on genuine image structure.

Watch

Extended reading notes

Core claim

The central discovery is that enforcing a bidirectional, cycle-consistent mapping between unlabeled images and ground-truth masks, rather than merely adversarially matching output distributions, pushes the segmentation network to preserve semantic content when labels are scarce. The model uses a generator $G_{IS}$ that maps images to masks and a generator $G_{SI}$ that maps masks back to images, trained with supervised cross-entropy on labeled pairs, LSGAN-style adversarial losses, a pixel-wise L1 cycle loss that reconstructs unlabeled images through mask space, and a cross-entropy cycle loss that reconstructs ground-truth masks through image space. An ablation on PASCAL VOC with 20% labeled data shows that each loss component contributes, with the mask-cycle loss making the largest single difference. The paper reports consistent mIoU gains over partial supervision, especially in the low-label regime, and says the method outperforms the adversarial semi-supervised segmentation approach used as comparison.

Load-bearing premise

The whole gain rests on the assumption that satisfying the pixel-level image reconstruction cycle forces the image-to-mask network to preserve the objects' semantic content, not just low-level statistics that are cheap to reconstruct.

Editorial extensions

If this is right

With only 10-20% labeled images, unlabeled images can improve mIoU by 2-4 points without any additional annotation effort.
The method does not require a domain shift between labeled and unlabeled data, a pre-trained generator, or paired image-mask examples, making it broadly applicable across dataset types.
The ablation shows that removing the label-cycle loss hurts more than removing the image-cycle loss, suggesting that future semi-supervised designs should emphasize consistency in mask space.
The benefit decreases as the labeled fraction grows, so the mechanism is most valuable in the low-label regime where annotation is the bottleneck.

Reading between the lines

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

The same cycle-consistent image-mask mapping could be adapted to weakly supervised settings, where image-level tags or scribbles replace full ground-truth masks in the label-cycle loss.
A testable extension is to replace the L1 image-cycle loss with a perceptual or feature-space cycle loss, which might prevent the segmenter from satisfying reconstruction through low-level texture statistics alone.
Because the framework does not require paired data or domain shift, it could plausibly transfer to video or volumetric medical data where dense annotations are especially scarce, though memory constraints would need to be addressed.

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.

Desk Editor's Note

read the letter

The new thing here is clean: applying CycleGAN-style cycle consistency directly between images and segmentation masks in the same domain, rather than across domains, as a regularizer for semi-supervised segmentation. That specific application is not in the earlier SSL or CycleGAN literature they cite, so the novelty claim holds. The paper also does several things well: it evaluates on three datasets with multiple labeled fractions, includes an ablation that separates cycle losses from GAN losses, and releases code.

The empirical pattern is mostly credible. On VOC and Cityscapes the proposed method beats the partial-supervision baseline at all label fractions, with the biggest margins at 10-20%, which matches the stated low-label regime. The ablation shows that removing either cycle loss hurts more than removing a discriminator loss, which is evidence the cycle terms are doing something beyond the GANs alone.

Now the soft spots. First, the abstract says "2-4% of improvement" and claims this broadly, but on ACDC the gains are below one point at every labeled fraction (e.g., 0.8418 to 0.8463 at 10%). The claim is true for VOC and Cityscapes but not for ACDC, so it overstates the evidence. Second, all results are single runs with no variance or seed information. For a paper whose central claim is an empirical gain of 2-4 points, that is a real weakness. Third, the only comparison method, Hung et al., is run without ImageNet pretraining, which the authors concede weakens it. That makes the "outperforms recent approaches" claim fragile.

The stress-test concern also lands. The image-cycle loss is an L1 reconstruction loss between a one-hot mask and a 3-channel image. Exact recovery is impossible, and the loss can be minimized by encoding photometric cues—color, texture, edges—that help GSI reconstruct the image without improving semantic accuracy. The label-cycle loss only uses labeled masks, so it cannot correct semantic errors on unlabeled images. The paper never measures whether the unsupervised objective is actually aligned with mIoU. This is not fatal: the method may still work as a regularizer via the GANs and cycle consistency in combination. But the mechanism is asserted, not demonstrated.

Citation pattern is fine—relevant CycleGAN, SSL, and segmentation works are covered, and there are no obvious missing references. This is a solid subfield contribution, not a conceptual breakthrough, and it deserves serious refereeing. We would need multi-seed results, error bars, a fairer comparison to Hung et al., and ideally an analysis of what the cycle loss actually learns before accepting. Recommend peer review.

Referee Report

4 major / 5 minor

Summary. The paper proposes a semi-supervised semantic segmentation method based on CycleGAN. In addition to the standard supervised cross-entropy loss on labeled images, the model trains two coupled generators (image-to-labels and labels-to-image), two pixel-wise discriminators, and two cycle-consistency losses, with the goal that the cycle-consistent mapping between unlabeled images and segmentation masks acts as an unsupervised regularizer. The method is evaluated on PASCAL VOC 2012, Cityscapes, and ACDC under 10%, 20%, 30%, and 50% labeled data, and compared against a partial-supervision baseline, a fully-supervised upper bound, and the adversarial SSL method of Hung et al. [11]. An ablation on VOC at 20% labeled data examines the contribution of each loss term.

Significance. If the reported gains are robust, the paper makes a useful empirical contribution: it shows that a CycleGAN-style cycle-consistent image-to-mask translation can improve semi-supervised segmentation, and it releases code and evaluates on three diverse benchmarks. The ablation study is a clear strength, and the discussion honestly acknowledges limitations such as reduced input resolution and the no-pretraining protocol for the comparison method. However, the significance is conditional: the headline 2-4% improvement claim is not supported on ACDC, the comparison with Hung et al. is weakened by a nonstandard protocol, and all results come from single runs without variance, so the magnitude and statistical reliability of the observed gains remain unclear.

major comments (4)

[Abstract and Section 4.3.1, Table 1] The abstract claims the model 'achieves 2-4% of improvement with respect to the baseline,' but Table 1 does not support this uniformly. Gains over the Partial baseline are 3.85 and 2.93 mIoU points on VOC at 10% and 20% labeled data, and 2.87 points on Cityscapes at 10%, but on ACDC the gains are only 0.45, 0.46, 0.19, and 0.27 points at 10%, 20%, 30%, and 50% respectively. The 2-4% statement should be revised to be dataset-specific or replaced by an honest aggregate, and the text should not imply that the magnitude is uniform across datasets.
[Section 4.3.1, comparison with Hung et al.] The comparison against Hung et al. [11] is not apples-to-apples because the authors train all methods from scratch, whereas the original method uses an ImageNet-pretrained generator. The paper concedes this could explain the poor result for [11], yet the abstract and Section 1 still claim the proposed method 'outperforms recent approaches for this task.' Since [11] is the only adversarial SSL baseline, this claim is not established under the original protocol. Please add the pretrained baseline or, at minimum, move the no-pretraining caveat into the results summary and temper the comparative claim accordingly.
[Section 4.3, all experiments] Every reported number comes from a single run with no standard deviation, multiple seeds, or significance test. This matters especially for the ACDC results, where the claimed improvements are below 0.5 mIoU points and could easily be within run-to-run noise. Without repeated runs, the strength of the central empirical claim—that cycle consistency gives a consistent boost in the low-label regime—cannot be assessed. Please provide at least 3-5 seeds with mean and standard deviation, or show that the differences are reproducible.
[Section 3.2, Eqs. (6)-(7) and Section 4.3.2, Table 2] The paper attributes the improvement to an 'unsupervised regularization effect' of cycle consistency, but no evidence is provided that the cycle losses are actually enforcing semantic correctness. The image cycle loss in Eq. (6) is a pixelwise L1 reconstruction loss; a segmentation network G_IS that encodes low-level photometric cues, such as colors or edges, could reduce this loss without improving mIoU. The label cycle loss in Eq. (7) is computed only on labeled masks Y_L, so it cannot directly correct semantic errors on unlabeled images, and the discriminator D_S only matches the marginal label distribution. The ablation in Table 2 shows that removing the cycle terms reduces mIoU, which is consistent with the claimed mechanism, but it is also consistent with the cycle losses acting as generic feature regularizers or GAN stabilizers. I would ask for an additional experiment or analysis that isolates the semantic effect, for example comparing against a reconstruction-only L1 regularizer, or reporting whether per-class errors on unlabeled images decrease in the directions encouraged by the label cycle.

minor comments (5)

[Section 4.3.2, text near Table 2] The text says 'w/o image cycle consistency loss (LS_cycle)' but the table uses LI_cycle; also, 'setting λ3, λ4, λ5 or λ6 to 0' refers to a non-existent λ6—the discriminator terms are weighted by λ4 and λ5.
[Section 4.1, ACDC paragraph] There is a duplicated word: 'well-deﬁned deﬁned pathologies' should read 'well-deﬁned pathologies'.
[Section 4.3.1, figure references] The text refers to 'Figures 7, 4 and 5' for the visual comparisons, but the corresponding figures appear to be Figures 3, 4, and 5. Please unify the figure numbering and in-text references.
[Section 3.2, Eq. (8)-(9)] The min-max formulation in Eq. (9) uses a total loss with negative discriminator terms, which is an unusual sign convention. A sentence clarifying how the alternating optimization maps onto the standard LSGAN objectives would help the reader.
[Section 3.3, implementation details] The text says 'implemented in Pytorch 3.3'; PyTorch versions are not typically numbered this way, and the capitalization should be PyTorch. Please verify the version and correct the wording.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the paper's central claim is an empirically tested regularization effect, not a derived prediction.

full rationale

The paper makes no derivation that reduces to its inputs. The proposed total loss in Eq. (8) is a standard combination of supervised cross-entropy (Eq. 1), a supervised image reconstruction loss (Eq. 3), two adversarial losses (Eqs. 4-5), and two cycle-consistency losses (Eqs. 6-7). The claimed outcome, improved mIoU under partial labeling, is evaluated by training on labeled subsets and measuring mean intersection-over-union on held-out validation sets (Table 1), so the reported numbers are not defined in terms of the training losses or fitted parameters. The hyperparameters lambda1-lambda5 are fixed manually and no parameter is fit to the test set and then reported as a prediction. The few self-citations (e.g., refs. [15] and [23] by co-authors) appear only in the related-work survey and do not carry the load of the method's justification. The skeptical concern that the cycle loss might reward low-level reconstructability rather than semantic accuracy is a hypothesis about the mechanism of an empirically observed regularizer; it is a correctness risk, not a circularity, because the empirical improvement is not assumed by construction. Accordingly, no circular step meets the evidentiary standard of quoting an equation or fitted quantity that is equivalent by definition to the claimed result.

Assumptions & free parameters 7 free parameters · 4 assumptions · 0 invented entities

The ledger is driven by hand-set hyperparameters and domain assumptions common in GAN-based SSL work. There are no new physical or mathematical entities. The most consequential assumptions are the same-domain condition and the semantic preservation of cycle consistency, neither of which is proven.

free parameters (7)

lambda_1 (image generator supervised loss weight) = 1.0
Set by hand in Section 3.3; no sensitivity analysis reported.
lambda_2 (labels cycle loss weight) = 0.05
Set by hand in Section 3.3; heavily downweights the label cycle loss compared with other terms.
lambda_3 (image cycle loss weight) = 1.0
Set by hand in Section 3.3; no sensitivity analysis reported.
lambda_4 (labels adversarial loss weight) = 1.0
Set by hand in Section 3.3; no sensitivity analysis reported.
lambda_5 (image adversarial loss weight) = 1.0
Set by hand in Section 3.3; no sensitivity analysis reported.
learning_rate = 2e-4
Initial learning rate with linear decay every 100 epochs, stated in Section 3.3.
batch_size = 5
Batch size used for all experiments, stated in Section 3.3.

assumptions (4)

domain assumption Labeled and unlabeled images are drawn from the same underlying distribution, with no domain shift.
The introduction explicitly says the setting has no domain shift between labeled and unlabeled data, which is required for the same-domain cycle regularization to transfer.
domain assumption The discriminator and generator training reaches a useful equilibrium without mode collapse or trivial solutions.
Eqs. (4) and (5) assume adversarial gradients are informative; the paper does not analyze failure modes such as the label discriminator being too strong or the segmentation generator collapsing to a constant mask.
domain assumption Cycle consistency losses preserve semantic content, not just low-level image statistics.
Eqs. (6) and (7) use pixel-level L1 and cross-entropy; the paper assumes these force G_IS to learn semantics rather than photometric shortcuts.
domain assumption The ground-truth masks YL are representative of all classes appearing in unlabeled and validation images at each label fraction.
The label discriminator is trained only on YL, so classes absent from the small labeled subset cannot be taught through adversarial or cycle losses.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Revisiting CycleGAN for semi-supervised segmentation." pith.science (2026). https://pith.science/paper/UHPLKZRK

@misc{pith2026190811569,
  author       = {Pith},
  title        = {Pith review of: Revisiting CycleGAN for semi-supervised segmentation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/UHPLKZRK}},
  note         = {Machine review of arXiv:1908.11569}
}

read the original abstract

In this work, we study the problem of training deep networks for semantic image segmentation using only a fraction of annotated images, which may significantly reduce human annotation efforts. Particularly, we propose a strategy that exploits the unpaired image style transfer capabilities of CycleGAN in semi-supervised segmentation. Unlike recent works using adversarial learning for semi-supervised segmentation, we enforce cycle consistency to learn a bidirectional mapping between unpaired images and segmentation masks. This adds an unsupervised regularization effect that boosts the segmentation performance when annotated data is limited. Experiments on three different public segmentation benchmarks (PASCAL VOC 2012, Cityscapes and ACDC) demonstrate the effectiveness of the proposed method. The proposed model achieves 2-4% of improvement with respect to the baseline and outperforms recent approaches for this task, particularly in low labeled data regime.

Figures

Figures reproduced from arXiv: 1908.11569 by the authors.

**Figure 2.** Examples of images, ground truth labels, generated images and generated labels obtained for three benchmark [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗

**Figure 3.** Visual comparisons on the PASCAL VOC 2012 dataset employing 20% of labeled images for training. [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗

Figures from the paper (4 more)

**Figure 4.** Figure 4: Visual comparisons on the Cityscape dataset employing 20% of labeled images for training. [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]

**Figure 5.** Figure 5: Visual comparisons on the ACDC dataset employing 20% of labeled images for training. [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]

**Figure 6.** Figure 6: Visual comparisons on the PASCAL VOC 2012 dataset employing 30% of labeled images for training. [PITH_FULL_IMAGE:figures/full_fig_p012_6.png]

**Figure 7.** Figure 7: Visual comparisons on the Cityscapes dataset employing 30% of labeled images for training. [PITH_FULL_IMAGE:figures/full_fig_p013_7.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

34 extracted references · 22 canonical work pages

[11]

Hung, Y .-H

W.-C. Hung, Y .-H. Tsai, Y .-T. Liou, Y .-Y . Lin, and M.-H. Yang. Adversarial learning for semi-supervised semantic segmentation. In Proceedings of the British Machine Vision Conference (BMVC), page 1, 2018

work page 2018
[1]

W. Bai, O. Oktay, M. Sinclair, H. Suzuki, M. Rajchl, G. Tar- roni, B. Glocker, A. King, P. M. Matthews, and D. Rueck- ert. Semi-supervised learning for network-based cardiac mr image segmentation. In International Conference on Med- ical Image Computing and Computer-Assisted Intervention , pages 253–260. Springer, 2017

work page 2017
[2]

C. Baur, S. Albarqouni, and N. Navab. Semi-supervised deep learning for fully convolutional networks. In International Conference on Medical Image Computing and Computer- Assisted Intervention, pages 311–319. Springer, 2017

work page 2017
[3]

Bernard, A

O. Bernard, A. Lalande, C. Zotti, F. Cervenansky, X. Yang, P.-A. Heng, I. Cetin, K. Lekadir, O. Camara, M. A. G. Ballester, et al. Deep learning techniques for automatic MRI cardiac multi-structures segmentation and diagnosis: Is the problem solved? IEEE Transactions on Medical Imaging , 2018

work page 2018
[4]

L. C. Chen, G. Papandreou, I. Kokkinos, K. Murphy, and A. L. Yuille. Deeplab: Semantic image segmentation with deep convolutional nets, atrous convolution, and fully con- nected crfs. IEEE Transactions on Pattern Analysis and Ma- chine Intelligence, 40(4):834–848, 2018

work page 2018
[5]

Cordts, M

M. Cordts, M. Omran, S. Ramos, T. Rehfeld, M. Enzweiler, R. Benenson, U. Franke, S. Roth, and B. Schiele. The cityscapes dataset for semantic urban scene understanding. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 3213–3223, 2016

2016
[6]

J. Dai, K. He, and J. Sun. Boxsup: Exploiting bounding boxes to supervise convolutional networks for semantic seg- mentation. In Proceedings of the IEEE International Con- ference on Computer Vision, pages 1635–1643, 2015

work page 2015
[7]

Everingham, L

M. Everingham, L. Van Gool, C. K. Williams, J. Winn, and A. Zisserman. The pascal visual object classes (voc) chal- lenge. International journal of computer vision, 88(2):303– 338, 2010

work page 2010

Show all 34 references

[8]

Goodfellow, J

I. Goodfellow, J. Pouget-Abadie, M. Mirza, B. Xu, D. Warde-Farley, S. Ozair, A. Courville, and Y . Bengio. Gen- erative adversarial nets. In Z. Ghahramani, M. Welling, C. Cortes, N. D. Lawrence, and K. Q. Weinberger, edi- tors, Advances in Neural Information Processing Systems...

2014
[9]

Hoffman, E

J. Hoffman, E. Tzeng, T. Park, J.-Y . Zhu, P. Isola, K. Saenko, A. A. Efros, and T. Darrell. Cycada: Cycle-consistent adver- sarial domain adaptation. arXiv preprint arXiv:1711.03213, 2017

2017 arXiv
[10]

Hoffman, D

J. Hoffman, D. Wang, F. Yu, and T. Darrell. FCNs in the wild: Pixel-level adversarial and constraint-based adapta- tion. arXiv preprint arXiv:1612.02649, 2016

2016 arXiv
[12]

Isola, J.-Y

P. Isola, J.-Y . Zhu, T. Zhou, and A. A. Efros. Image-to- image translation with conditional adversarial networks. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 1125–1134, 2017. Image Ground truth Full Partial Hung et al. [11] Ours Ima...

2017
[13]

Jiang, Y .-C

J. Jiang, Y .-C. Hu, N. Tyagi, P. Zhang, A. Rimner, G. S. Mageras, J. O. Deasy, and H. Veeraraghavan. Tumor-aware, adversarial domain adaptation from ct to mri for lung cancer segmentation. In International Conference on Medical Im- age Computing and Computer-Assisted Interven...

2018
[14]

Johnson, A

J. Johnson, A. Alahi, and L. Fei-Fei. Perceptual losses for real-time style transfer and super-resolution. In European conference on computer vision , pages 694–711. Springer, 2016

2016
[15]

Kervadec, J

H. Kervadec, J. Dolz, M. Tang, E. Granger, Y . Boykov, and I. B. Ayed. Constrained-CNN losses for weakly supervised segmentation. Medical image analysis, 2019

2019
[16]

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

2014 arXiv
[17]

D. Lin, J. Dai, J. Jia, K. He, and J. Sun. Scribble- sup: Scribble-supervised convolutional networks for seman- tic segmentation. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition , pages 3159– 3167, 2016

2016
[18]

J. Long, E. Shelhamer, and T. Darrell. Fully convolutional networks for semantic segmentation. In Proceedings of the IEEE conference on computer vision and pattern recogni- tion, pages 3431–3440, 2015

2015
[19]

X. Mao, Q. Li, H. Xie, R. Y . Lau, Z. Wang, and S. Paul Smol- ley. Least squares generative adversarial networks. In Pro- ceedings of the IEEE International Conference on Computer Vision, pages 2794–2802, 2017

2017
[20]

Min and X

S. Min and X. Chen. A robust deep attention network to noisy labels in semi-supervised biomedical segmentation. arXiv preprint arXiv:1807.11719, 2018

2018 arXiv
[21]

Papandreou, L.-C

G. Papandreou, L.-C. Chen, K. Murphy, and A. L. Yuille. Weakly-and semi-supervised learning of a DCNN for seman- tic image segmentation. In ICCV, 2015

2015
[22]

Paszke, S

A. Paszke, S. Gross, S. Chintala, G. Chanan, E. Yang, Z. De- Vito, Z. Lin, A. Desmaison, L. Antiga, and A. Lerer. Auto- matic differentiation in PyTorch. InNIPS Autodiff Workshop, 2017

2017
[23]

J. Peng, G. Estradab, M. Pedersoli, and C. Desrosiers. Deep co-training for semi-supervised image segmentation. arXiv preprint arXiv:1903.11233, 2019

1903 arXiv
[24]

C. S. Perone, P. Ballester, R. C. Barros, and J. Cohen- Adad. Unsupervised domain adaptation for medical imag- ing segmentation with self-ensembling. arXiv preprint arXiv:1811.06042, 2018

2018 arXiv
[25]

Rajchl, M

M. Rajchl, M. C. Lee, O. Oktay, K. Kamnitsas, J. Passerat- Palmbach, W. Bai, M. Damodaram, M. A. Rutherford, J. V . Hajnal, B. Kainz, et al. Deepcut: Object segmentation from bounding box annotations using convolutional neural net- works. IEEE transactions on medical imaging ,...

2017
[26]

Ronneberger, P

O. Ronneberger, P. Fischer, and T. Brox. U-net: Convo- lutional networks for biomedical image segmentation. In International Conference on Medical image computing and computer-assisted intervention , pages 234–241. Springer, 2015

2015
[27]

Souly, C

N. Souly, C. Spampinato, and M. Shah. Semi supervised se- mantic segmentation using generative adversarial network. In Computer Vision (ICCV), 2017 IEEE International Con- ference on, pages 5689–5697. IEEE, 2017

2017
[28]

Tsai, W.-C

Y .-H. Tsai, W.-C. Hung, S. Schulter, K. Sohn, M.-H. Yang, and M. Chandraker. Learning to adapt structured output space for semantic segmentation. In Computer Vision and Pattern Recognition (CVPR), 2018

2018
[29]

Tzeng, J

E. Tzeng, J. Hoffman, K. Saenko, and T. Darrell. Adversarial discriminative domain adaptation. In Computer Vision and Pattern Recognition (CVPR), volume 1, page 4, 2017

2017
[30]

Ulyanov, A

D. Ulyanov, A. Vedaldi, and V . Lempitsky. Instance normal- ization: The missing ingredient for fast stylization. arXiv preprint arXiv:1607.08022, 2016

2016 arXiv
[31]

Zhang, L

Y . Zhang, L. Yang, J. Chen, M. Fredericksen, D. P. Hughes, and D. Z. Chen. Deep adversarial networks for biomed- ical image segmentation utilizing unannotated images. In International Conference on Medical Image Computing and Computer-Assisted Intervention, pages 408–416, 2017

2017
[32]

Y . Zhou, X. He, L. Huang, L. Liu, F. Zhu, S. Cui, and L. Shao. Collaborative learning of semi-supervised segmen- tation and classiﬁcation for medical images. In Proceed- ings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 2079–2088, 2019

2019
[33]

Y . Zhou, Y . Wang, P. Tang, W. Shen, E. K. Fishman, and A. L. Yuille. Semi-supervised multi-organ segmentation via multi-planar co-training. arXiv preprint arXiv:1804.02586, 2018

2018 arXiv
[34]

J.-Y . Zhu, T. Park, P. Isola, and A. A. Efros. Unpaired image- to-image translation using cycle-consistent adversarial net- works. In Proceedings of the IEEE International Conference on Computer Vision, pages 2223–2232, 2017. Partial Ours Class 10% 20% 30% 10% 20% 30% 1 0.423...

2017

Pith tools

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

[1] [11]

Hung, Y .-H

W.-C. Hung, Y .-H. Tsai, Y .-T. Liou, Y .-Y . Lin, and M.-H. Yang. Adversarial learning for semi-supervised semantic segmentation. In Proceedings of the British Machine Vision Conference (BMVC), page 1, 2018

work page 2018

[2] [1]

W. Bai, O. Oktay, M. Sinclair, H. Suzuki, M. Rajchl, G. Tar- roni, B. Glocker, A. King, P. M. Matthews, and D. Rueck- ert. Semi-supervised learning for network-based cardiac mr image segmentation. In International Conference on Med- ical Image Computing and Computer-Assisted Intervention , pages 253–260. Springer, 2017

work page 2017

[3] [2]

C. Baur, S. Albarqouni, and N. Navab. Semi-supervised deep learning for fully convolutional networks. In International Conference on Medical Image Computing and Computer- Assisted Intervention, pages 311–319. Springer, 2017

work page 2017

[4] [3]

Bernard, A

O. Bernard, A. Lalande, C. Zotti, F. Cervenansky, X. Yang, P.-A. Heng, I. Cetin, K. Lekadir, O. Camara, M. A. G. Ballester, et al. Deep learning techniques for automatic MRI cardiac multi-structures segmentation and diagnosis: Is the problem solved? IEEE Transactions on Medical Imaging , 2018

work page 2018

[5] [4]

L. C. Chen, G. Papandreou, I. Kokkinos, K. Murphy, and A. L. Yuille. Deeplab: Semantic image segmentation with deep convolutional nets, atrous convolution, and fully con- nected crfs. IEEE Transactions on Pattern Analysis and Ma- chine Intelligence, 40(4):834–848, 2018

work page 2018

[6] [5]

Cordts, M

M. Cordts, M. Omran, S. Ramos, T. Rehfeld, M. Enzweiler, R. Benenson, U. Franke, S. Roth, and B. Schiele. The cityscapes dataset for semantic urban scene understanding. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 3213–3223, 2016

2016

[7] [6]

J. Dai, K. He, and J. Sun. Boxsup: Exploiting bounding boxes to supervise convolutional networks for semantic seg- mentation. In Proceedings of the IEEE International Con- ference on Computer Vision, pages 1635–1643, 2015

work page 2015

[8] [7]

Everingham, L

M. Everingham, L. Van Gool, C. K. Williams, J. Winn, and A. Zisserman. The pascal visual object classes (voc) chal- lenge. International journal of computer vision, 88(2):303– 338, 2010

work page 2010