Pith. sign in

REVIEW 5 major objections 5 minor 1 cited by

CSG: A Context-Semantic Guided Diffusion Approach in De Novo Musculoskeletal Ultrasound Image Generation

T0 review · 5 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read This paper introduces CSG, a dual-conditioning diffusion model that generates musculoskeletal ultrasound images from a semantic mask and a style-matched context image, and reports that experts cannot reliably distinguish the synthetic…

desk verdict A useful dual-conditioning method for ultrasound synthesis whose quantitative claims outrun the evidence because context selection can leak near-duplicate real frames into the evaluation. read the letter →

arxiv 2412.05833 v1 pith:IM5ZQCYL submitted 2024-12-08 cs.CV eess.IV

Elay Dahan , Hedda Cohen Indelman , Angeles M. Perez-Agosto , Carmit Shiran , Gopal Avinash , Doron Shaked , Nati Daniel This is my paper · ORCID
classification cs.CVeess.IV
keywords musculoskeletalultrasoundlatentdiffusionmodelsdualconditioningsemanticmaskssyntheticmedicalimagessegmentationTuringtestdataaugmentation
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

CSG is a method for generating synthetic musculoskeletal (MSK) ultrasound images by giving a latent diffusion model two independent controls: a semantic mask that fixes the anatomical structure and a context image that provides the texture style. The context image is chosen automatically as the dataset image with the closest neural style features to the prompt. The paper's claim is that this dual conditioning produces de novo images that are closer to real ultrasound images than a mask-to-image translation baseline, that adding these images to a segmentation training set improves Dice scores for all classes and for the rare discontinuity-in-tendon-fiber finding, and that clinical experts cannot reliably tell the synthetic images from real ones in a Turing test. If the claim holds, CSG offers a controllable, scalable route to enriching ultrasound datasets with rare pathological findings.

What carries the argument

The load-bearing object is a dual-conditioned latent diffusion model with modified classifier-free guidance. The sampling score estimate is $\tilde{e}_\theta(z_t,c_S,c_C)=\phi+s_S(e_\theta(z_t,c_S,\emptyset)-\phi)+s_C(e_\theta(z_t,c_S,c_C)-e_\theta(z_t,c_S,\emptyset))$, where $c_S$ is the semantic mask latent, $c_C$ is the context image latent, and the guidance scales are $s_S=1.5$, $s_C=2.5$. The semantic mask controls geometry; the context image, selected by MSE over neural style features following the method of [30], controls texture. The mask generator is a StyleGAN fine-tuned on MSK masks, and variability extensions use an LLM to translate text prompts into classical image operations plus Poisson image editing for texture blending.

What would settle it

Take the same trained CSG model and regenerate every test image twice: once with its original style-matched context and once with a context image forced to come from a different patient or a different ultrasound clip, then compare FID and the Turing test results. If the synthetic outputs degrade substantially or become near-duplicates of the held-out context image, the reported realism comes from context leakage rather than from a learned coupling of mask and texture.

Watch

Extended reading notes

Core claim

The central discovery is that semantic and contextual conditioning can be combined in a single latent diffusion sampler without sacrificing either geometric faithfulness or textural realism. The paper shows this through a three-fold validation: downstream semantic segmentation improves by 26% overall and 23% on the DITF finding when CSG images are added to the real training data; the synthetic images have FID 7.816 versus 24.189 for the baseline; and in a Turing test run by trained personnel and clinical experts, CSG images were the least distinguishable from real images. The paper also reports that the CSG extension, which generates semantic masks with a fine-tuned StyleGAN and edits geometry and texture through text-guided operations and Poisson blending, can expand the variability space of generated anatomies and textures.

Load-bearing premise

The method assumes that the automatically selected style context, which is frequently taken from the same ultrasound clip as the prompt image, supplies texture guidance without leaking the identity of the target image into the synthetic output.

Editorial extensions

If this is right

  • Adding CSG synthetic images to real training data raises mean Dice for the segmentation model from 0.46 to 0.58 across all classes and from 0.39 to 0.48 on the DITF finding.
  • CSG synthetic images sit closer to the real-image distribution than the baseline translation model, with FID 7.816 versus 24.189.
  • Trained personnel and clinical experts could not reliably distinguish CSG images from real ultrasound images in a three-set Turing test.
  • The mask-generation and editing extensions let a user prescribe geometry changes by text and blend textures from another image, so rare anatomical variants can be synthesized on demand.

Reading between the lines

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

  • A stricter patient-level split in which the context image is never drawn from the same patient or clip as the prompt would test whether the reported fidelity survives genuine out-of-patient generation; the paper does not report such a split.
  • Because the context selection is based only on style features, CSG could in principle be applied to other ultrasound anatomies or imaging modalities whenever paired masks and images are available.
  • The text-to-computer-vision mask editing pipeline implies a user interface for clinicians to specify augmentation, but its downstream effect on diagnostic accuracy beyond segmentation is not yet measured.
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

5 major / 5 minor

Summary. The paper introduces CSG, a dual-conditioning pipeline for generating musculoskeletal (MSK) ultrasound images. A semantic mask controls anatomy and a context image controls texture; the context for each training image is selected as the nearest neighbor in a style-feature space. The generation stage fine-tunes a latent diffusion model with a modified classifier-free guidance rule (Eq. (1)), and scaling is proposed via StyleGAN-based mask generation and text/geometry editing. Validation is threefold: downstream semantic segmentation (Table 2), image-quality metrics (Table 3), and a visual Turing test (Section 3.3). The reported results show improved segmentation Dice, lower FID/KLD/KST relative to a single baseline, and human difficulty in distinguishing CSG outputs from real images.

Significance. If the central claims hold, CSG would be a useful controllable data-generation method for an ultrasound domain where rare pathological findings are scarce; the dual semantic/context conditioning and the variability-extension modules are practically relevant ideas. The paper also makes a specific, falsifiable claim: synthetic images improve a downstream segmentation model beyond a control and beyond one baseline. However, the validation as presented does not yet establish the 'de novo' and 'indistinguishability' claims because the context-selection procedure may leak near-duplicate information, and the quantitative evidence lacks error bars, significance testing, and a second baseline. The manuscript is a reasonable candidate after substantial experimental strengthening, not in its current form.

major comments (5)
  1. [Section 2.1.1 and Fig. 2] The context-selection procedure is load-bearing for the 'de novo' and high-fidelity claims, but it may allow near-duplication: for each prompt image, the context is the most style-similar non-equal image in the dataset, and the paper explicitly states that these are 'often from the same ultrasound clip'. Because the model is trained to map the prompt's semantic mask plus this near-identical context back to the prompt's appearance, the low FID and high human confusion could reflect copying of real frames rather than synthesis of novel anatomy/texture combinations. Please quantify the context-target similarity (e.g., pixel or feature distances, fraction of pairs from the same clip/patient), and evaluate generation with contexts constrained to hold-out patients or to clips different from the mask source. Without such an experiment, the claim that CSG generates 'de novo' and 'undistinguishable' images is not supported.
  2. [Section 3.3] The Turing-test reporting is too ambiguous to support the abstract's claim that CSG images are 'undistinguishable from real images according to a Turing test'. The paper reports 'mean TPR of 0.2967, mean PPV of 0.24 and F1 of 0.238' but does not define the positive class or provide the full confusion matrix, making it impossible to assess whether these numbers indicate that humans failed to detect synthetic images. The inter-observer agreement values (82.6%, 68%, 57%, 62.6%) are not mapped to the three set compositions described in the text. Please report the full confusion matrix for each composition, define TPR/PPV with respect to real versus synthetic labels, and state the human accuracy and chance level.
  3. [Section 3.1 and Table 2] The segmentation improvement is presented without any measure of variance: there are no error bars, confidence intervals, or significance tests, and the table appears to report a single run. The text states an improvement 'by a factor of 26 and 23', but the numbers 0.58 versus 0.46 and 0.48 versus 0.39 correspond to relative improvements of about 26% and 23%, not factors of 26 and 23; this wording should be corrected. Please report results over multiple seeds and include statistical testing, and ideally compare with at least one additional generative baseline beyond the self-cited [27].
  4. [Section 2.3.1 versus Section 3.1] There is an inconsistency in the reported cohort size: Section 2.3.1 describes a dataset of 388 ultrasound images from 90 patients, while Section 3.1 states '388 real images from 124 subjects'. Since the segmentation and quality comparisons depend on the dataset, this discrepancy must be resolved and the correct patient count stated consistently.
  5. [Section 2.2.1] The mask-generation extension is described as 'unlimited and unbiased', but the filtering mechanism that 'exclude[s] generated masks that lack significant pathologies' is not specified. This filtering biases the generated distribution toward pathology-present examples and should be quantified; otherwise the unbiasedness claim is unsupported. At minimum, report the fraction of masks rejected and the criteria for 'significant pathologies'.
minor comments (5)
  1. [Section 2.1.2, Eq. (1)] The notation in Eq. (1) is inconsistent: the text uses 'CC, CS' and 'CC = ∅ or CS = ∅', but the equation and surrounding text use c_C and c_S. Please unify the notation.
  2. [Table 1] Several percentages in Table 1 appear to be off by a factor of 100: Calcification is listed as 0.11% but 44/388 is 11.3%, and Bone irregularity is listed as 0.31% but 121/388 is 31.2%. Please correct these values and clarify the denominator.
  3. [Section 2.3.1] The sentence 'All images are resized to a constant resolution of 1535x796 and 1044x646 pixels' lists two resolutions without explaining which images use which; please clarify.
  4. [Section 3.3] The phrase 'de novo' appears both hyphenated and unhyphenated across the manuscript; please standardize the spelling.
  5. [Section 3.4] The variability-extension results are shown only qualitatively in Fig. 5; a quantitative evaluation (e.g., segmentation consistency before and after mask editing, or texture realism scores for Poisson-edited images) would make the extension claims more concrete.

Circularity Check

1 steps flagged · score 2.0 of 10

Mild circularity in the realism evaluation: the context input is chosen as the nearest real image to the target, so FID and Turing results are partly forced by input selection rather than by fully de novo synthesis.

  1. other [Section 2.1.1, Fig. 2; evaluated in Sections 3.2 and 3.3]
    "Then, for each image (denoted as prompt image) in the dataset, we find the most similar non-equal feature vector by computing the Mean Squared Error (MSE) between these two features. By that, we were able to find a visually similar image for any given image in the dataset. We then constructed a dataset of two input images of semantic mask and context image paired with the ground-truth ultrasound image. ... Indeed, the selected images are often from the same ultrasound clip (e.g. top row)."

    The context condition is, by construction, the nearest real image to the very image the model is trained to reproduce. Since the model learns to map the semantic mask plus this near-duplicate context back to the ground-truth ultrasound, the similarity metrics and Turing test can be satisfied by copying the input context's texture and structure rather than by generating novel anatomy or appearance. The reported low FID and high human confusion are therefore not an independent confirmation of de novo synthesis; they are partly guaranteed by the input-selection rule. The paper does demonstrate out-of-dataset contexts, so this is a partial, not total, circularity.

full rationale

The central derivation of CSG is otherwise self-contained: it fine-tunes a pretrained latent diffusion model with a semantic mask and a context image as dual conditions, using the modified classifier-free-guidance score in Eq. (1). No parameter is fitted to the evaluation metrics, and the downstream segmentation improvement is an external benchmark rather than a restatement of the training objective. The self-citations ([27], [33], [34], and [4]) are used as baselines, mask-generation pointers, or related work; they are not load-bearing, and no uniqueness theorem or ansatz is imported from them. The only circularity is the context-selection step: because the context is the nearest real image to the prompt image, generation quality measured by FID and by human indistinguishability can be achieved by recalling or copying the context input. This weakens the strength of the 'de novo' claim, but since the paper also shows generation from out-of-dataset contexts and the main method does not reduce to its evaluation metrics, the overall circularity is mild.

Assumptions & free parameters 2 free parameters · 5 assumptions · 0 invented entities

The central method rests on the assumption that an anatomical mask plus a texture reference fully determines a realistic ultrasound image, and that VGG style features are a good proxy for ultrasound texture. It also assumes transfer from histopathology-pretrained StyleGAN to MSK masks. No new physical entities are introduced; two guidance scales are hand-set.

free parameters (2)
  • sS (semantic guidance scale) = 1.5
    Set manually in Section 2.1.2 and used for all experiments; affects the strength of mask conditioning.
  • sC (context guidance scale) = 2.5
    Set manually in Section 2.1.2; affects the strength of texture conditioning.
assumptions (5)
  • domain assumption Paired semantic masks and context images are sufficient conditioning to synthesize realistic ultrasound images.
    Section 2.1.2 assumes the dual-conditioned LDM can map mask plus context to a real-looking image; no clinical generative model is provided.
  • domain assumption VGG style features, trained on natural images, order ultrasound textures by visual similarity.
    Section 2.1.1 uses style encodings from a pre-trained CNN (Gatys et al.) to select context images; this transfer is assumed.
  • ad hoc to paper StyleGAN pretrained on BRECAHAD histopathology can be fine-tuned to generate MSK masks.
    Section 2.2.1 initializes from BRECAHAD weights; the suitability of this initialization for MSK anatomy is not demonstrated independently.
  • domain assumption High-fidelity synthetic images lie near real images in the Inception embedding space.
    Section 3.2 invokes the manifold hypothesis to justify using FID, KST, and KLD as quality proxies.
  • standard math The latent diffusion and classifier-free guidance formulations of [23,32] are correct.
    The method relies on the standard equations from these references without re-deriving them.

how reviews work

0 comments
Cite this review

Pith. "Pith review of CSG: A Context-Semantic Guided Diffusion Approach in De Novo Musculoskeletal Ultrasound Image Generation." pith.science (2026). https://pith.science/paper/IM5ZQCYL

@misc{pith2026241205833,
  author       = {Pith},
  title        = {Pith review of: CSG: A Context-Semantic Guided Diffusion Approach in De Novo Musculoskeletal Ultrasound Image Generation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/IM5ZQCYL}},
  note         = {Machine review of arXiv:2412.05833}
}
read the original abstract

The use of synthetic images in medical imaging Artificial Intelligence (AI) solutions has been shown to be beneficial in addressing the limited availability of diverse, unbiased, and representative data. Despite the extensive use of synthetic image generation methods, controlling the semantics variability and context details remains challenging, limiting their effectiveness in producing diverse and representative medical image datasets. In this work, we introduce a scalable semantic and context-conditioned generative model, coined CSG (Context-Semantic Guidance). This dual conditioning approach allows for comprehensive control over both structure and appearance, advancing the synthesis of realistic and diverse ultrasound images. We demonstrate the ability of CSG to generate findings (pathological anomalies) in musculoskeletal (MSK) ultrasound images. Moreover, we test the quality of the synthetic images using a three-fold validation protocol. The results show that the synthetic images generated by CSG improve the performance of semantic segmentation models, exhibit enhanced similarity to real images compared to the baseline methods, and are undistinguishable from real images according to a Turing test. Furthermore, we demonstrate an extension of the CSG that allows enhancing the variability space of images by synthetically generating augmentations of anatomical geometries and textures.

Figures

Figures reproduced from arXiv: 2412.05833 by the authors.

Figure 1
Figure 1. (a) Depiction of the differences between Image Translation GANs and Vanilla GANs, where the former [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Examples of a query and contextual similar image pairs. On the left: a query image. On the right: The most [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Example synthetic ultrasound images generated by our CSG method. A semantic mask (on the left) and [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Turing test results of the MSK synthetic images and a comparison of generated image examples. [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 5
Figure 5. Figure 5: Visualization of our extending the variability space. Our method allows for controlling the generation of the [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

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

  1. Prompt Mechanisms in Medical Imaging: A Comprehensive Survey

    eess.IV 2025-06 conditional novelty 4.0 of 10

    A broad survey that organizes prompt mechanisms for medical image generation, segmentation, and classification into a two-dimensional taxonomy of core technologies and clinical applications.

Reference graph

Works this paper leans on

49 extracted references · 33 canonical work pages · cited by 1 Pith paper

  1. [27]

    Between Generating Noise and Generating Images: Noise in the Correct Frequency Improves the Quality of Synthetic Histopathology Images for Digital Pathology

    N. Daniel, E. Aknin, A. Larey, Y . Peretz, G. Sela, Y . Fisher, and Y . Savir, “Between generating noise and generating images: Noise in the correct frequency improves the quality of synthetic histopathology images for digital pathology,”arXiv preprint arXiv:2302.06549, 2023

    work page Pith review arXiv 2023

  2. [1]

    Development and clinical application of artificial intelligence assistant system for rotator cuff ultrasound scanning,

    R. Tang, Z. Li, L. Jiang, J. Jiang, B. Zhao, L. Cui, G. Zhou, X. Chen, and D. Jiang, “Development and clinical application of artificial intelligence assistant system for rotator cuff ultrasound scanning,” Ultrasound in Medicine & Biology, vol. 50, no. 2, pp. 251–257, 2024

    work page 2024

  3. [2]

    Muscle cross-sectional area segmentation in transverse ultrasound images using vision transformers,

    S. Katakis, N. Barotsis, A. Kakotaritis, P. Tsiganos, G. Economou, E. Panagiotopoulos, and G. Panayiotakis, “Muscle cross-sectional area segmentation in transverse ultrasound images using vision transformers,”Diagnostics, vol. 13, no. 2, p. 217, 2023

    work page 2023

  4. [3]

    Deep learning segmentation of transverse musculoskeletal ultrasound images for neuromuscular disease assessment,

    F. Marzola, N. van Alfen, J. Doorduin, and K. M. Meiburger, “Deep learning segmentation of transverse musculoskeletal ultrasound images for neuromuscular disease assessment,” Computers in Biology and Medicine, vol. 135, p. 104623, 2021

    work page 2021

  5. [4]

    Semantic Segmentation Refiner for Ultrasound Applications with Zero-Shot Foundation Models

    H. C. Indelman, E. Dahan, A. M. Perez-Agosto, C. Shiran, D. Shaked, and N. Daniel, “Semantic segmentation refiner for ultrasound applications with zero-shot foundation models,” 2024. [Online]. Available: https:// arxiv.org/abs/2404.16325

    work page Pith review arXiv 2024

  6. [5]

    Deep learning with convolutional neural network in radiology,

    K. Yasaka, H. Akai, A. Kunimatsu, S. Kiryu, and O. Abe, “Deep learning with convolutional neural network in radiology,”Japanese journal of radiology, vol. 36, pp. 257–272, 2018

    work page 2018

  7. [6]

    Deep learning: a primer for radiologists,

    G. Chartrand, P. M. Cheng, E. V orontsov, M. Drozdzal, S. Turcotte, C. J. Pal, S. Kadoury, and A. Tang, “Deep learning: a primer for radiologists,” Radiographics, vol. 37, no. 7, pp. 2113–2131, 2017

    work page 2017

  8. [7]

    Overview of deep learning in medical imaging,

    K. Suzuki, “Overview of deep learning in medical imaging,” Radiological physics and technology, vol. 10, no. 3, pp. 257–273, 2017

    work page 2017

Show all 49 references
  1. [8]

    Machine learning approach for biopsy-based identification of eosinophilic esophagitis reveals importance of global features,

    T. Czyzewski, N. Daniel, M. Rochman, J. M. Caldwell, G. A. Osswald, M. H. Collins, M. E. Rothenberg, and Y . Savir, “Machine learning approach for biopsy-based identification of eosinophilic esophagitis reveals importance of global features,” IEEE open journal of engineering i...

    2021

  2. [9]

    A deep multi-label segmentation network for eosinophilic esophagitis whole slide biopsy diagnostics,

    N. Daniel, A. Larey, E. Aknin, G. A. Osswald, J. M. Caldwell, M. Rochman, M. H. Collins, G.-Y . Yang, N. C. Arva, K. E. Capocelli, M. E. Rothenberg, and Y . Savir, “A deep multi-label segmentation network for eosinophilic esophagitis whole slide biopsy diagnostics,” in 2022 44...

    2022

  3. [10]

    Harnessing artificial intelligence to infer novel spatial biomarkers for the diagnosis of eosinophilic esophagitis,

    A. Larey, E. Aknin, N. Daniel, G. A. Osswald, J. M. Caldwell, M. Rochman, T. Wasserman, M. H. Collins, N. C. Arva, G.-Y . Yang, M. E. Rothenberg, and Y . Savir, “Harnessing artificial intelligence to infer novel spatial biomarkers for the diagnosis of eosinophilic esophagitis,...

    2022

  4. [11]

    Translational ai and deep learning in diagnostic pathology,

    A. Serag, A. Ion-Margineanu, H. Qureshi, R. McMillan, M.-J. Saint Martin, J. Diamond, P. O’Reilly, and P. Hamilton, “Translational ai and deep learning in diagnostic pathology,”Frontiers in medicine, vol. 6, p. 185, 2019

    2019

  5. [12]

    Artificial intelligence and digital pathology: challenges and opportunities,

    H. R. Tizhoosh and L. Pantanowitz, “Artificial intelligence and digital pathology: challenges and opportunities,” Journal of pathology informatics, vol. 9, no. 1, p. 38, 2018

    2018

  6. [13]

    Sketch guided and progressive growing gan for realistic and editable ultrasound image synthesis,

    J. Liang, X. Yang, Y . Huang, H. Li, S. He, X. Hu, Z. Chen, W. Xue, J. Cheng, and D. Ni, “Sketch guided and progressive growing gan for realistic and editable ultrasound image synthesis,” Medical Image Analysis, vol. 79, p. 102461, 2022

    2022

  7. [14]

    A multi-constraint hybrid network for ultrasound image synthesis,

    K. Cao, D. Ai, D. Xiao, and J. Yang, “A multi-constraint hybrid network for ultrasound image synthesis,” in2023 IEEE 7th Information Technology and Mechatronics Engineering Conference (ITOEC), vol. 7. IEEE, 2023, pp. 2147–2151

    2023

  8. [15]

    Realistic ultrasound image synthesis for improved classification of liver disease,

    H. Che, S. Ramanathan, D. J. Foran, J. L. Nosher, V . M. Patel, and I. Hacihaliloglu, “Realistic ultrasound image synthesis for improved classification of liver disease,” inSimplifying Medical Ultrasound: Second International Workshop, ASMUS 2021, Held in Conjunction with MICC...

    2021

  9. [16]

    Self-supervised ultrasound to mri fetal brain image synthesis,

    J. Jiao, A. I. Namburete, A. T. Papageorghiou, and J. A. Noble, “Self-supervised ultrasound to mri fetal brain image synthesis,” IEEE Transactions on Medical Imaging, vol. 39, no. 12, pp. 4413–4424, 2020

    2020

  10. [17]

    Resvit: Residual vision transformers for multimodal medical image synthesis,

    O. Dalmaz, M. Yurt, and T. Çukur, “Resvit: Residual vision transformers for multimodal medical image synthesis,” IEEE Transactions on Medical Imaging, vol. 41, no. 10, pp. 2598–2614, 2022

    2022

  11. [18]

    Generative adversarial networks,

    I. Goodfellow, J. Pouget-Abadie, M. Mirza, B. Xu, D. Warde-Farley, S. Ozair, A. Courville, and Y . Bengio, “Generative adversarial networks,”Communications of the ACM, vol. 63, no. 11, pp. 139–144, 2020. 9 CSG: A Context-Semantic Guided Diffusion Approach DAHAN E ET AL

    2020

  12. [19]

    Unsupervised representation learning with deep convolutional generative adversarial networks,

    A. Radford, L. Metz, and S. Chintala, “Unsupervised representation learning with deep convolutional generative adversarial networks,” arXiv preprint arXiv:1511.06434, 2015

    2015 arXiv

  13. [20]

    Synthetic medical images from dual generative adversarial networks,

    J. T. Guibas, T. S. Virdi, and P. S. Li, “Synthetic medical images from dual generative adversarial networks,”arXiv preprint arXiv:1709.01872, 2017

    2017 arXiv

  14. [21]

    Conditional generative adversarial nets,

    M. Mirza and S. Osindero, “Conditional generative adversarial nets,” arXiv preprint arXiv:1411.1784, 2014

    2014 arXiv

  15. [22]

    Deep unsupervised learning using nonequilib- rium thermodynamics,

    J. Sohl-Dickstein, E. Weiss, N. Maheswaranathan, and S. Ganguli, “Deep unsupervised learning using nonequilib- rium thermodynamics,” in International conference on machine learning. PMLR, 2015, pp. 2256–2265

    2015

  16. [23]

    High-resolution image synthesis with latent diffusion models,

    R. Rombach, A. Blattmann, D. Lorenz, P. Esser, and B. Ommer, “High-resolution image synthesis with latent diffusion models,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2022, pp. 10 684–10 695

    2022

  17. [24]

    Auto-Encoding Variational Bayes,

    D. P. Kingma and M. Welling, “Auto-Encoding Variational Bayes,” in2nd International Conference on Learning Representations, ICLR, Conference Track Proceedings, 2014

    2014

  18. [25]

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

    J.-Y . Zhu, T. Park, P. Isola, and A. A. Efros, “Unpaired image-to-image translation using cycle-consistent adversarial networks,” in Proceedings of the IEEE international conference on computer vision, 2017, pp. 2223– 2232

    2017

  19. [26]

    High-resolution image synthesis and semantic manipulation with conditional gans,

    T.-C. Wang, M.-Y . Liu, J.-Y . Zhu, A. Tao, J. Kautz, and B. Catanzaro, “High-resolution image synthesis and semantic manipulation with conditional gans,” in Proceedings of the IEEE conference on computer vision and pattern recognition, 2018, pp. 8798–8807

    2018

  20. [28]

    Medgan: Medical image translation using gans,

    K. Armanious, C. Jiang, M. Fischer, T. Küstner, T. Hepp, K. Nikolaou, S. Gatidis, and B. Yang, “Medgan: Medical image translation using gans,” Computerized medical imaging and graphics, vol. 79, p. 101684, 2020

    2020

  21. [29]

    Generative adversarial network in medical imaging: A review,

    X. Yi, E. Walia, and P. Babyn, “Generative adversarial network in medical imaging: A review,”Medical image analysis, vol. 58, p. 101552, 2019

    2019

  22. [30]

    A neural algorithm of artistic style,

    L. A. Gatys, A. S. Ecker, and M. Bethge, “A neural algorithm of artistic style,”arXiv preprint arXiv:1508.06576, 2015

    2015 arXiv

  23. [31]

    U-net: Convolutional networks for biomedical image segmentation,

    O. Ronneberger, P. Fischer, and T. Brox, “U-net: Convolutional networks for biomedical image segmentation,” in Medical Image Computing and Computer-Assisted Intervention – MICCAI 2015, N. Navab, J. Hornegger, W. M. Wells, and A. F. Frangi, Eds. Cham: Springer International Pub...

    2015

  24. [32]

    Instructpix2pix: Learning to follow image editing instructions,

    T. Brooks, A. Holynski, and A. A. Efros, “Instructpix2pix: Learning to follow image editing instructions,” in CVPR, 2023

    2023

  25. [33]

    Depas: De-novo pathology semantic masks using a generative model,

    A. Larey, N. Daniel, E. Aknin, Y . Fisher, and Y . Savir, “Depas: De-novo pathology semantic masks using a generative model,” arXiv preprint arXiv:2302.06513, 2023

    2023 arXiv

  26. [34]

    Priorpath: Coarse-to-fine approach for controlled de-novo pathology semantic masks generation,

    N. Daniel, M. Nathan, E. Azeroual, Y . Fisher, and Y . Savir, “Priorpath: Coarse-to-fine approach for controlled de-novo pathology semantic masks generation,” arXiv preprint arXiv:2411.16515, 2024

    2024 arXiv

  27. [35]

    Training generative adversarial networks with limited data,

    T. Karras, M. Aittala, J. Hellsten, S. Laine, J. Lehtinen, and T. Aila, “Training generative adversarial networks with limited data,” Advances in neural information processing systems, vol. 33, pp. 12 104–12 114, 2020

    2020

  28. [36]

    Brecahad: a dataset for breast cancer histopathological annotation and diagnosis,

    A. Aksac, D. J. Demetrick, T. Ozyer, and R. Alhajj, “Brecahad: a dataset for breast cancer histopathological annotation and diagnosis,” BMC research notes, vol. 12, no. 1, pp. 1–3, 2019

    2019

  29. [37]

    Llama: Open and efficient foundation language models,

    H. Touvron, T. Lavril, G. Izacard, X. Martinet, M.-A. Lachaux, T. Lacroix, B. Rozière, N. Goyal, E. Hambro, F. Azhar, A. Rodriguez, A. Joulin, E. Grave, and G. Lample, “Llama: Open and efficient foundation language models,” 2023. [Online]. Available: https://arxiv.org/abs/2302.13971

    2023 arXiv

  30. [38]

    Palm: Scaling language modeling with pathways,

    A. Chowdhery, S. Narang, J. Devlin, M. Bosma, G. Mishra, A. Roberts, P. Barham, H. W. Chung, C. Sutton, S. Gehrmann et al., “Palm: Scaling language modeling with pathways,” Journal of Machine Learning Research, vol. 24, no. 240, pp. 1–113, 2023

    2023

  31. [39]

    Poisson image editing,

    P. Pérez, M. Gangnet, and A. Blake, “Poisson image editing,” inSeminal Graphics Papers: Pushing the Boundaries, Volume 2, 2023, pp. 577–582

    2023

  32. [40]

    Laion-400m: Open dataset of clip-filtered 400 million image-text pairs,

    C. Schuhmann, R. Vencu, R. Beaumont, R. Kaczmarczyk, C. Mullis, A. Katta, T. Coombes, J. Jitsev, and A. Komatsuzaki, “Laion-400m: Open dataset of clip-filtered 400 million image-text pairs,” arXiv preprint arXiv:2111.02114, 2021. 10 CSG: A Context-Semantic Guided Diffusion App...

    2021 arXiv

  33. [41]

    Decoupled weight decay regularization,

    I. Loshchilov and F. Hutter, “Decoupled weight decay regularization,” arXiv preprint arXiv:1711.05101, 2017

    2017 arXiv

  34. [42]

    Pytorch: An imperative style, high-performance deep learning library,

    A. Paszke, S. Gross, F. Massa, A. Lerer, J. Bradbury, G. Chanan, T. Killeen, Z. Lin, N. Gimelshein, L. Antiga et al., “Pytorch: An imperative style, high-performance deep learning library,”Advances in neural information processing systems, vol. 32, 2019

    2019

  35. [43]

    Evaluating kolmogorov’s distribution,

    G. Marsaglia, W. W. Tsang, and J. Wang, “Evaluating kolmogorov’s distribution,”Journal of statistical software, vol. 8, pp. 1–4, 2003

    2003

  36. [44]

    A. W. Bowman and A. Azzalini, Applied smoothing techniques for data analysis: the kernel approach with S-Plus illustrations. OUP Oxford, 1997, vol. 18

    1997

  37. [45]

    Gans trained by a two time-scale update rule converge to a local nash equilibrium,

    M. Heusel, H. Ramsauer, T. Unterthiner, B. Nessler, and S. Hochreiter, “Gans trained by a two time-scale update rule converge to a local nash equilibrium,”Advances in neural information processing systems, vol. 30, 2017

    2017

  38. [46]

    Rethinking the inception architecture for computer vision,

    C. Szegedy, V . Vanhoucke, S. Ioffe, J. Shlens, and Z. Wojna, “Rethinking the inception architecture for computer vision,” in Proceedings of the IEEE conference on computer vision and pattern recognition, 2016, pp. 2818–2826

    2016

  39. [47]

    Evaluating white matter lesion segmentations with refined sørensen-dice analysis,

    A. Carass, S. Roy, A. Gherman, J. Reinhold, A. Jesson, T. Arbel, O. Maier, H. Handels, M. Ghafoorian, B. Platel, A. Birenbaum, H. Greenspan, D. Pham, C. Crainiceanu, P. Calabresi, J. Prince, W. Roncal, R. Shinohara, and I. Oguz, “Evaluating white matter lesion segmentations wi...

    2020

  40. [48]

    Fast r-cnn,

    R. Girshick, “Fast r-cnn,” in Proceedings of the IEEE international conference on computer vision , 2015, pp. 1440–1448

    2015

  41. [49]

    Rethinking atrous convolution for semantic image segmentation,

    L.-C. Chen, G. Papandreou, F. Schroff, and H. Adam, “Rethinking atrous convolution for semantic image segmentation,” ArXiv, vol. abs/1706.05587, 2017. 11

    2017 arXiv

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.