Pith. sign in

REVIEW 4 major objections 6 minor 36 references

IMPROVE: Improving Medical Plausibility without Reliance on HumanValidation -- An Enhanced Prototype-Guided Diffusion Framework

T0 review · 4 major / 6 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read The paper claims that seeding a diffusion model's class embeddings with learned prototype vectors raises the share of expert-approved synthetic medical images from about 48% to 60–63% on two datasets, without any human feedback during…

desk verdict A sensible but incremental prototype-guided diffusion tweak; the downstream F1 gains look real, but the headline plausibility claim rests on a single unblinded rater per dataset, no statistics, and two unexplained missing classes. read the letter →

arxiv 2411.17535 v1 pith:RKP3Q5I2 submitted 2024-11-26 cs.CV

Anurag Shandilya , Swapnil Bhat , Akshat Gautam , Subhash Yadav , Siddharth Bhatt , Deval Mehta , Kshitij Jadhav This is my paper · ORCID
classification cs.CV
keywords diffusionmodelsmedicalimagegenerationprototypelearningbiologicalplausibilityclassifier-freeguidancesyntheticdataaugmentationbonemarrowcytologydermatoscopy
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's central claim is that medical plausibility of synthetic images, not pixel fidelity, is the bottleneck in medical image generation, and that this bottleneck can be addressed without the costly human-feedback loop used in prior RLHF-based work. The proposed fix is to condition a diffusion model on learned per-class prototype vectors—compact class summaries distilled from image embeddings—rather than on randomly initialized class embeddings. On bone marrow cytology and dermatoscopic skin-lesion datasets, expert raters judged roughly 63% and 60% of the prototype-guided images plausible, versus about 48% for classifier-free guidance. Downstream classifiers trained on the synthetic images improved by about 11 and 4 F1 points, narrowing the gap to classifiers trained on real images. If the evaluation holds, this gives a cheaper route to medically plausible synthetic data for augmenting rare and long-tailed medical datasets.

What carries the argument

The central object is the per-class prototype vector: a learned representative embedding for each class, computed from CLIP image embeddings using distance-based cross-entropy and a prototype loss that rewards intra-class compactness and inter-class separability. This vector is concatenated with the sinusoidal time embedding and injected into the UNet layers of the diffusion model, with its weights frozen during training. It carries the argument because it replaces the randomly initialized class embedding of classifier-free guidance with a semantically meaningful target, steering generation toward class-typical morphology without requiring a reward model or human ratings.

What would settle it

Have two or more independent experts, blinded to which model produced each image, rate the same generated sets using the paper's stated criteria; if inter-rater agreement is low, or if the prototype-guided advantage over classifier-free guidance disappears under blinded scoring, the central claim is unsupported.

Watch

Extended reading notes

Core claim

The paper's discovery is that the conditioning signal in a diffusion model can carry medical semantics: replacing random class embeddings with frozen prototype vectors learned from CLIP embeddings of 100 images per class makes the denoising process land in biologically plausible regions of image space. With identical architecture and hyperparameters, the only intervention is the prototype-initialized, frozen class embedding concatenated with the time embedding and fed into the UNet's layers. The paper reports that the fraction of expert-approved images rises from 48.8% to 60% on HAM10000 and from 48.07% to 62.57% on Bone Marrow, with gains in most individual classes and particularly large gains for hard cell types like hairy cells and basophils. No expert feedback is used during training; the experts appear only in the final evaluation. The same synthetic datasets also train better downstream classifiers, giving F1 scores of 74.58% and 75.08% versus 63.40% and 71.10% for the classifier-free baseline, although still below the 79.66% and 85.91% achieved with real images.

Load-bearing premise

The load-bearing premise is that the single expert's plausibility judgments are a valid and reproducible gold standard: one pathologist and one dermatologist rated unblinded image sets, with no inter-rater reliability or statistical testing reported.

Editorial extensions

If this is right

  • Synthetic datasets produced this way can augment rare and long-tailed medical classes with a higher fraction of expert-acceptable images than classifier-free diffusion, at no extra annotation cost during training.
  • The method transfers across at least two modalities (hematology and dermatology), suggesting the prototype-initialization mechanism is not tied to one type of tissue or stain.
  • Training a downstream classifier on prototype-guided synthetic images yields F1 gains of about 11 points (Bone Marrow) and 4 points (HAM10000) over the classifier-free baseline, so the plausibility improvement is accompanied by usable class signal.
  • Classifiers trained on real images still outperform those trained on synthetic images, so the method's promise is augmentation rather than full replacement of real data.
  • Because prototypes are learned from a small sample (100 images per class) and then frozen, the pipeline is computationally light enough for small medical datasets.

Reading between the lines

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

  • A direct comparison against the RLHF-aligned generator of prior work, at matched annotation budgets, would test whether prototype guidance captures most of the plausibility benefit without the expert-in-the-loop cost; the paper only compares against classifier-free guidance.
  • The same recipe—learn compact per-class prototypes and freeze them as diffusion conditioning—could be applied to other structured generation domains, such as pathology patches, retinal imaging, or even non-medical domains where class-typical structure matters.
  • Because each dataset was scored by one unblinded expert, a plausible expectation is that the effect size will shrink under multi-rater blinded protocols; that expectation is directly testable.
  • The dependency on CLIP embeddings means the method's transfer to modalities where CLIP features are weak may require a domain-specific feature extractor; this is an open question, not a result of the paper.
Share X Bluesky LinkedIn Reddit HN

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 proposes IMPROVE, a prototype-guided diffusion framework for medical image generation. The authors learn per-class prototypes from CLIP embeddings of the training images, then freeze these prototypes and use them in place of randomly initialized class embeddings to condition a UNet-based diffusion model, with the prototype and time embeddings concatenated and fed into the network. Experiments are performed on the Bone Marrow and HAM10000 datasets at 64x64 resolution, with 100 images per class used for training. A single pathologist (Bone Marrow) and a single dermatologist (HAM10000) rate generated images for medical plausibility, and downstream ResNext-50 classifiers are trained on the synthetic images. The paper reports plausibility increases from 48.8% to 60% for HAM10000 and from 48.07% to 62.57% for Bone Marrow, and downstream F1 improvements of about 11 and 4 percentage points over a classifier-free guidance baseline.

Significance. If the reported gains are real and reproducible, the method would be a meaningful contribution: it removes the need for human feedback during training, which is a major bottleneck in medical image generation, and it shows a downstream benefit in classification. The idea of using learned prototypes as a substitute for human guidance is interesting and builds naturally on existing prototype-learning and diffusion literature. The paper also provides per-class plausibility breakdowns and an independent downstream evaluation, which are both strengths. However, the current evaluation is not sufficient to establish the central claim: the plausibility measurements rely on single, unblinded raters with no statistical analysis, and two Bone Marrow classes are omitted from the plausibility table without justification. The method description is also underspecified regarding the baseline and guidance mechanism. The potential significance is real, but the evidence as presented is suggestive rather than conclusive.

major comments (4)
  1. [Section 5.3, Tables 3 and 4] The central claim of improved plausibility rests entirely on ratings from a single expert per dataset (one pathologist for Bone Marrow, one dermatologist for HAM10000), with no blinding, no randomization of presentation order, no inter-rater reliability, and no statistical testing. The reported increases of 11.2 and 14.5 percentage points are well within the range that an unblinded rater's expectation could produce, especially when the images are 64x64 patches rated against qualitative criteria. To support the headline claim, the authors should provide ratings from multiple blinded experts, report inter-rater agreement, and compute confidence intervals or a significance test for the differences. Without this, the evidence is anecdotal.
  2. [Table 4 and Section 6] Table 4 reports plausibility for only 14 of the 16 Bone Marrow classes listed in Table 1; the classes 'ART (Artefact)' and 'NIF (Not identifiable)' are omitted. The paper does not state whether the average plausibility (48.07% to 62.57%) is computed over the 14 listed classes or over all 16, and no justification is given for the omission. The downstream classifier in Table 5 is similarly trained on 14 cell types. This omission could affect the claimed average improvement and the conclusion that the improvement is consistent across two domains. Please clarify the exact classes used in each analysis and report results for all classes or explain why these two are excluded.
  3. [Section 4.2 and Section 5.3] The description of the baseline and the proposed guidance mechanism is underspecified. The paper refers to the baseline as 'Classifier Free Guidance', but standard classifier-free guidance requires randomly dropping the conditioning information during training and using a guidance scale at sampling. The text does not state whether this dropout was applied, nor does it give the guidance scale for either model. The proposed method is described only as concatenating frozen prototype embeddings with time embeddings and feeding them to UNet layers; no equation or algorithmic pseudocode defines how the prototype conditioning modifies the noise prediction. Because the claimed advantage over the baseline could in principle be due to differences in conditioning handling rather than the prototypes themselves, the comparison needs a precise specification of both models and identical training and sampling protocols.
  4. [Tables 5 and 6] The downstream classification results are reported as single-point precision, recall, and F1 values without variance or significance testing. For example, Table 5 reports F1 of 74.58% for the proposed method versus 63.40% for the baseline, but no confidence intervals are given; given that only one synthetic training set of each type was generated and one classifier trained, the observed gap could be within run-to-run variability. Please report results over multiple synthetic datasets and training seeds, with mean plus/minus standard deviation and a significance test, to support the claim that the improvements are consistent.
minor comments (6)
  1. [Section 6] In the sentence describing the downstream classifier, 'seven cell types for HAM1000 dataset' should read 'HAM10000 dataset'.
  2. [Table 3] The header 'Morphological Cell Type' is inaccurate for the dermatology dataset; rename it to 'Lesion Type' or similar.
  3. [Section 5.3] The annotation platform is written 'Labelstudio'; the standard name is 'Label Studio'.
  4. [References] Reference [31] for HAM10000 lists only the dataset name as author; the full citation (Tschandl, Rosendahl, and Kittler, Scientific Data, 2018) should be provided. Reference [17] is also missing author names.
  5. [Section 4.2] The architecture name 'UNet' is written as 'UNET' inconsistently; please standardize.
  6. [Abstract] The abstract says the method improves plausibility 'without the need for any human feedback,' but the evaluation uses human experts; consider clarifying that human feedback is not used during training or generation, only for evaluation.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the plausibility improvement is measured against external expert ratings and downstream classifiers, not derived from the method's own objective.

full rationale

The paper's central claim is that prototype-guided diffusion generates medical images with higher expert-rated plausibility and better downstream classification performance than a classifier-free guidance baseline. This claim is evaluated empirically: plausible-image fractions are provided by one pathologist and one dermatologist, and downstream utility is measured by training a ResNext-50 classifier on synthetic images and testing on a held-out set of real images. These are external benchmarks, not quantities reconstructed from the method's training objective. The prototypes are learned from the same training images that train the diffusion model, but that is standard training rather than definitional circularity: the diffusion loss is noise prediction, and the expert plausibility labels are not used in training. The baseline uses randomly initialized class embeddings with the same hyperparameters, so the reported gains are empirical differences, not identities. Citations to prior work, including ProtoDiffusion [3], CPL [35], and the RLHF medical-image work [30], are external published methods and datasets, not self-citations by the present authors, and none of these citations is used to define away the evaluation. The weakness of relying on a single unblinded expert per dataset is a validity and reproducibility concern, but it is not a circularity of the derivation chain. No equation or fitted parameter is shown to reduce the reported plausibility improvement to an input, so the paper is not circular.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The central claim rests on CLIP embeddings being informative for medical classes, on the prototype learning procedure being effective with 100 images per class, and on the human expert evaluation being a valid metric. No new physical entities are introduced.

free parameters (3)
  • Prototype learning hyperparameters (number of prototypes per class, gamma, lambda, optimizer) = not specified
    Section 4.1 says prototypes are learned with 'a different classifier' but gives no algorithm or values; these choices determine prototype quality and the downstream result.
  • CLIP model variant = not specified (embedding dimension 768 implies ViT-L/14)
    Section 4.1 states a pretrained CLIP model with 768-dimensional embeddings, but the exact variant is not named; different CLIP models produce different embeddings and prototype quality.
  • Diffusion training hyperparameters (noise schedule, dropout, etc.) = partially specified (lr 1e-4, batch 16, 1500 epochs, channel configuration)
    Section 5.3 lists some values but not the full schedule; these affect generation quality and the fairness of the comparison with the baseline.
assumptions (4)
  • domain assumption CLIP embeddings trained on natural images transfer to medical image classes
    Section 4.1 uses a pretrained CLIP model to embed bone marrow and dermoscopy images; the method assumes these embeddings capture medically relevant class structure.
  • domain assumption Prototypes learned from 100 images per class are representative of each class
    Section 5.3 uses 100 images per class for both prototype learning and diffusion training; if this sample is not representative, the prototypes are unreliable.
  • domain assumption Single expert judgment per dataset is a valid gold standard for medical plausibility
    Section 5.3 relies on one pathologist and one dermatologist; no inter-rater reliability or blinding is reported, so the evaluation may not be reproducible.
  • standard math Standard diffusion and prototype-learning equations (Eqs. 1-14) hold as derived in cited works
    The method builds on DDPM and CPL formulas from prior literature without re-derivation.

how reviews work

0 comments
Cite this review

Pith. "Pith review of IMPROVE: Improving Medical Plausibility without Reliance on HumanValidation -- An Enhanced Prototype-Guided Diffusion Framework." pith.science (2026). https://pith.science/paper/RKP3Q5I2

@misc{pith2026241117535,
  author       = {Pith},
  title        = {Pith review of: IMPROVE: Improving Medical Plausibility without Reliance on HumanValidation -- An Enhanced Prototype-Guided Diffusion Framework},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/RKP3Q5I2}},
  note         = {Machine review of arXiv:2411.17535}
}
read the original abstract

Generative models have proven to be very effective in generating synthetic medical images and find applications in downstream tasks such as enhancing rare disease datasets, long-tailed dataset augmentation, and scaling machine learning algorithms. For medical applications, the synthetically generated medical images by such models are still reasonable in quality when evaluated based on traditional metrics such as FID score, precision, and recall. However, these metrics fail to capture the medical/biological plausibility of the generated images. Human expert feedback has been used to get biological plausibility which demonstrates that these generated images have very low plausibility. Recently, the research community has further integrated this human feedback through Reinforcement Learning from Human Feedback(RLHF), which generates more medically plausible images. However, incorporating human feedback is a costly and slow process. In this work, we propose a novel approach to improve the medical plausibility of generated images without the need for human feedback. We introduce IMPROVE:Improving Medical Plausibility without Reliance on Human Validation - An Enhanced Prototype-Guided Diffusion Framework, a prototype-guided diffusion process for medical image generation and show that it substantially enhances the biological plausibility of the generated medical images without the need for any human feedback. We perform experiments on Bone Marrow and HAM10000 datasets and show that medical accuracy can be substantially increased without human feedback.

Figures

Figures reproduced from arXiv: 2411.17535 by the authors.

Figure 1
Figure 1. Prototype generation represents the first step of the train [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Time embeddings and learnt prototypes are concatenated and passed at various stages in the UNET model to guide the diffusion [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 4
Figure 4. Generated images for 5 classes of Bone Marrow Dataset [PITH_FULL_IMAGE:figures/full_fig_p006_4.png] view at source ↗
Figures from the paper (3 more)
Figure 3
Figure 3. Figure 3: Generated images for 5 classes of Bone Marrow Dataset [PITH_FULL_IMAGE:figures/full_fig_p006_3.png]
Figure 5
Figure 5. Figure 5: Generated images for 4 classes of Skin Cancer Dataset [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 6
Figure 6. Figure 6: Generated images for 4 classes of Skin Cancer Dataset [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

36 extracted references · 18 canonical work pages

  1. [1]

    What does dall-e 2 know about radiology? Journal of Medical Internet Research, 25:e43110, 2023

    Lisa C Adams, Felix Busch, Daniel Truhn, Marcus R Makowski, Hugo JWL Aerts, and Keno K Bressem. What does dall-e 2 know about radiology? Journal of Medical Internet Research, 25:e43110, 2023. 1

    work page 2023

  2. [2]

    Spot the fake lungs: Generating synthetic medical images using neural dif- fusion models

    Hazrat Ali, Shafaq Murad, and Zubair Shah. Spot the fake lungs: Generating synthetic medical images using neural dif- fusion models. In Irish Conference on Artificial Intelligence and Cognitive Science, pages 32–39. Springer, 2022. 1, 2

    work page 2022

  3. [3]

    Protodiffusion: classifier-free diffusion guid- ance with prototype learning

    Gulcin Baykal, Halil Faruk Karagoz, Taha Binhuraib, and Gozde Unal. Protodiffusion: classifier-free diffusion guid- ance with prototype learning. In Asian Conference on Ma- chine Learning, pages 106–120. PMLR, 2024. 3, 4

    work page 2024

  4. [4]

    Retrieval-augmented diffusion models

    Andreas Blattmann, Robin Rombach, Kaan Oktay, Jonas M¨uller, and Bj ¨orn Ommer. Retrieval-augmented diffusion models. Advances in Neural Information Processing Sys- tems, 35:15309–15324, 2022. 2

    2022

  5. [5]

    A survey on active learning and human-in-the-loop deep learn- ing for medical image analysis

    Samuel Budd, Emma C Robinson, and Bernhard Kainz. A survey on active learning and human-in-the-loop deep learn- ing for medical image analysis. Medical image analysis , 71:102062, 2021. 1

    work page 2021

  6. [6]

    Adapting pretrained vision-language foundational models to medical imaging domains

    Pierre Chambon, Christian Bluethgen, Curtis P Langlotz, and Akshay Chaudhari. Adapting pretrained vision-language foundational models to medical imaging domains. arXiv preprint arXiv:2210.04133, 2022. 1

    arXiv 2022

  7. [7]

    Synthetic data in machine learning for medicine and healthcare.Nature Biomedical En- gineering, 5(6):493–497, 2021

    Richard J Chen, Ming Y Lu, Tiffany Y Chen, Drew FK Williamson, and Faisal Mahmood. Synthetic data in machine learning for medicine and healthcare.Nature Biomedical En- gineering, 5(6):493–497, 2021. 1

    work page 2021

  8. [8]

    Diffusion models beat gans on image synthesis

    Prafulla Dhariwal and Alexander Nichol. Diffusion models beat gans on image synthesis. Advances in neural informa- tion processing systems, 34:8780–8794, 2021. 2, 3

    2021

Show all 36 references
  1. [9]

    Protodiff: learning to learn prototypical networks by task- guided diffusion

    Yingjun Du, Zehao Xiao, Shengcai Liao, and Cees Snoek. Protodiff: learning to learn prototypical networks by task- guided diffusion. Advances in Neural Information Process- ing Systems, 36:46304–46322, 2023. 2

    2023

  2. [10]

    Taming transformers for high-resolution image synthesis

    Patrick Esser, Robin Rombach, and Bjorn Ommer. Taming transformers for high-resolution image synthesis. In Pro- ceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 12873–12883, 2021. 2

    2021

  3. [11]

    Deepncm: Deep nearest class mean classifiers

    Samantha Guerriero, Barbara Caputo, and Thomas Mensink. Deepncm: Deep nearest class mean classifiers. 2018. 2

    2018

  4. [12]

    Gans trained by a two time-scale update rule converge to a local nash equilib- rium

    Martin Heusel, Hubert Ramsauer, Thomas Unterthiner, Bernhard Nessler, and Sepp Hochreiter. Gans trained by a two time-scale update rule converge to a local nash equilib- rium. Advances in neural information processing systems , 30, 2017. 2

    2017

  5. [13]

    Denoising dif- fusion probabilistic models

    Jonathan Ho, Ajay Jain, and Pieter Abbeel. Denoising dif- fusion probabilistic models. Advances in neural information processing systems, 33:6840–6851, 2020. 3

    2020

  6. [14]

    Cascaded diffu- sion models for high fidelity image generation

    Jonathan Ho, Chitwan Saharia, William Chan, David J Fleet, Mohammad Norouzi, and Tim Salimans. Cascaded diffu- sion models for high fidelity image generation. Journal of Machine Learning Research, 23(47):1–33, 2022. 2

    2022

  7. [15]

    Classifier-free diffusion guidance

    Jonathan Ho and Tim Salimans. Classifier-free diffusion guidance. arXiv preprint arXiv:2207.12598, 2022. 3

    2022 arXiv

  8. [16]

    Variational prototyping-encoder: One-shot learning with prototypical images

    Junsik Kim, Tae-Hyun Oh, Seokju Lee, Fei Pan, and In So Kweon. Variational prototyping-encoder: One-shot learning with prototypical images. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages 9462–9470, 2019. 2

    2019

  9. [17]

    An expert-annotated dataset of bone marrow cytology in hematologic malignan- cies, 2021

    Christian Matek, Stephan Krappe, Christian M ¨unzenmayer, Torsten Haferlach, and Carsten Marr. An expert-annotated dataset of bone marrow cytology in hematologic malignan- cies, 2021. 5

    2021

  10. [18]

    A multimodal comparison of latent denois- ing diffusion probabilistic models and generative adversarial networks for medical image synthesis

    Gustav M ¨uller-Franzes, Jan Moritz Niehues, Firas Khader, Soroosh Tayebi Arasteh, Christoph Haarburger, Christiane Kuhl, Tianci Wang, Tianyu Han, Teresa Nolte, Sven Nebelung, et al. A multimodal comparison of latent denois- ing diffusion probabilistic models and generative ad...

    2023

  11. [19]

    Generation of anonymous chest radiographs using la- tent diffusion models for training thoracic abnormality clas- sification systems

    Kai Packh ¨auser, Lukas Folle, Florian Thamm, and Andreas Maier. Generation of anonymous chest radiographs using la- tent diffusion models for training thoracic abnormality clas- sification systems. In 2023 IEEE 20th International Sympo- sium on Biomedical Imaging (ISBI), page...

    2023

  12. [20]

    Brain imaging generation with latent diffusion models

    Walter HL Pinaya, Petru-Daniel Tudosiu, Jessica Dafflon, Pedro F Da Costa, Virginia Fernandez, Parashkev Nachev, Sebastien Ourselin, and M Jorge Cardoso. Brain imaging generation with latent diffusion models. In MICCAI Work- shop on Deep Generative Models, pages 117–126. Springer,

  13. [21]

    Learning transferable visual models from natural language supervi- sion

    Alec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh, Gabriel Goh, Sandhini Agarwal, Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, et al. Learning transferable visual models from natural language supervi- sion. In International conference on machine learning, ...

    2021

  14. [22]

    Deep learning for medical image processing: Overview, challenges and the future

    Muhammad Imran Razzak, Saeeda Naz, and Ahmad Zaib. Deep learning for medical image processing: Overview, challenges and the future. Classification in BioApps: Au- tomation of decision making, pages 323–350, 2018. 1

    2018

  15. [23]

    Assessing generative models via precision and recall

    Mehdi SM Sajjadi, Olivier Bachem, Mario Lucic, Olivier Bousquet, and Sylvain Gelly. Assessing generative models via precision and recall. Advances in neural information pro- cessing systems, 31, 2018. 2

    2018

  16. [24]

    Progressive distillation for fast sampling of diffusion models

    Tim Salimans and Jonathan Ho. Progressive distillation for fast sampling of diffusion models. arXiv preprint arXiv:2202.00512, 2022. 2

    2022 arXiv

  17. [25]

    Laion-5b: An open large-scale dataset for training next generation image-text models

    Christoph Schuhmann, Romain Beaumont, Richard Vencu, Cade Gordon, Ross Wightman, Mehdi Cherti, Theo Coombes, Aarush Katta, Clayton Mullis, Mitchell Worts- man, et al. Laion-5b: An open large-scale dataset for training next generation image-text models. Advances in Neural In- f...

    2022

  18. [26]

    P-odn: Prototype-based open deep net- work for open set recognition

    Yu Shu, Yemin Shi, Yaowei Wang, Tiejun Huang, and Yonghong Tian. P-odn: Prototype-based open deep net- work for open set recognition. Scientific reports, 10(1):7146,

  19. [27]

    Prototypical networks for few-shot learning

    Jake Snell, Kevin Swersky, and Richard Zemel. Prototypical networks for few-shot learning. Advances in neural informa- tion processing systems, 30, 2017. 2

    2017

  20. [28]

    Deep unsupervised learning using nonequilibrium thermodynamics

    Jascha Sohl-Dickstein, Eric Weiss, Niru Maheswaranathan, and Surya Ganguli. Deep unsupervised learning using nonequilibrium thermodynamics. In International confer- ence on machine learning, pages 2256–2265. PMLR, 2015. 2

    2015

  21. [29]

    Denoising diffusion implicit models

    Jiaming Song, Chenlin Meng, and Stefano Ermon. Denoising diffusion implicit models. arXiv preprint arXiv:2010.02502, 2020. 2, 5

    2010 arXiv

  22. [30]

    Aligning synthetic medical images with clinical knowledge using human feedback

    Shenghuan Sun, Greg Goldgof, Atul Butte, and Ahmed M Alaa. Aligning synthetic medical images with clinical knowledge using human feedback. Advances in Neural In- formation Processing Systems, 36, 2024. 1, 2

    2024

  23. [31]

    The HAM10000 dataset, a large collec- tion of multi-source dermatoscopic images of common pig- mented skin lesions, 2018

    Philipp Tschandl. The HAM10000 dataset, a large collec- tion of multi-source dermatoscopic images of common pig- mented skin lesions, 2018. 5

    2018

  24. [32]

    The evolution of video quality measurement: From psnr to hybrid metrics

    Stefan Winkler and Praveen Mohandas. The evolution of video quality measurement: From psnr to hybrid metrics. IEEE transactions on Broadcasting , 54(3):660–668, 2008. 2

    2008

  25. [33]

    Medical long-tailed learning for imbalanced data: bibliometric analy- sis

    Zheng Wu, Kehua Guo, Entao Luo, Tian Wang, Shoujin Wang, Yi Yang, Xiangyuan Zhu, and Rui Ding. Medical long-tailed learning for imbalanced data: bibliometric analy- sis. Computer Methods and Programs in Biomedicine, page 108106, 2024. 1

    2024

  26. [34]

    Unsupervised feature learning via non-parametric instance discrimination

    Zhirong Wu, Yuanjun Xiong, Stella X Yu, and Dahua Lin. Unsupervised feature learning via non-parametric instance discrimination. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 3733–3742,

  27. [35]

    Robust classification with convolutional proto- type learning

    Hong-Ming Yang, Xu-Yao Zhang, Fei Yin, and Cheng- Lin Liu. Robust classification with convolutional proto- type learning. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 3474–3482,

  28. [36]

    Rethinking semantic segmentation: A proto- type view

    Tianfei Zhou, Wenguan Wang, Ender Konukoglu, and Luc Van Gool. Rethinking semantic segmentation: A proto- type view. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages 2582– 2593, 2022. 2

    2022

Pith tools

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