Pith. sign in

REVIEW 4 major objections 5 minor 41 references

Representation-driven Endoscopic Visual Embedding Alignment for Latent Generation

T0 review · 4 major / 5 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read REVEAL is a generative foundation model for endoscopy trained on 4.82M endoscopic frames that produces high-fidelity images and features that beat endoscopy-specific encoders in several benchmarks.

desk verdict Credible first large-scale generative foundation model for endoscopy with systematic ablations, but the headline comparisons rest on an untested patient-level dedup assumption that needs explicit clarification. read the letter →

arxiv 2608.07176 v1 pith:JLR24LBU submitted 2026-08-07 cs.CV cs.AI

Francisco Caetano , Tim J.M. Jaspers , Haiko Middeljans , Martijn R. Jong , Rixta A.H. van Eijck van Heslinga , Floor Slooter , Albert J. de Groof , Jacques J. Bergman
show 2 more authors
Peter H.N. De With Fons van der Sommen
This is my paper · ORCID
classification cs.CVcs.AI
keywords endoscopygenerativefoundationmodelrepresentationalignmentflowmatchingdiffusiontransformerimageinpaintingoutpaintinglinearprobing
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

REVEAL is a generative foundation model for endoscopy trained on GastroNet-5M (GN-5M), a multicenter collection of 4,820,653 unlabeled endoscopic frames. The paper argues that representation alignment with an encoder pretrained on the same endoscopic distribution—rather than a general-purpose natural-image encoder—is what makes synthesis faithful, and that the resulting frozen transformer is itself a competitive visual feature extractor. On linear-probing benchmarks for polyp and Barrett's esophagus neoplasia, REVEAL's layer-8 features exceed EndoViT and Endo-FM, and its FID on generated GN-5M samples reaches 5.32 after 250k iterations with the SiT-L/2 backbone. If true, large-scale generative pretraining offers a viable alternative path to clinical visual representations, alongside masked-image-modeling encoders, with a single backbone for synthesis, editing, and downstream classification.

What carries the argument

The load-bearing mechanism is the representation-alignment objective in its iREPA variant: during training the student's hidden patch tokens $h_t$ are pushed toward the teacher's patch embeddings $y^*$ by maximizing cosine similarity, $$ \mathcal{L}_{\text{REPA}} = -\mathbb{E}_{x,\epsilon,t}\left[\frac{1}{N}\sum_{n=1}^N \operatorname{sim}(y^*[n], h_\$\varphi$(h_t[n]))\right], $$ added to the denoising loss with weight $\lambda=1$. Two details make it work for endoscopy: a convolutional projection head that keeps neighboring tokens spatially coherent, and an instance-style spatial normalization applied to the teacher features that suppresses global variance and boosts local anatomical contrast. The teacher is a DINOv3 ViT-B/16 checkpoint continued on GN-5M, and the student is a SiT flow-matching transformer operating on a frozen SD2 VAE latent space. This mechanism injects domain-specific patch-level structure into the generative trajectory while leaving the trained transformer usable as a feature extractor.

What would settle it

Compute FID on a held-out endoscopic image set from hospitals that did not contribute to GN-5M and run a blinded clinical reader test in which experts try to distinguish REVEAL samples from real frames; if FID degrades sharply or experts reliably spot synthesis artifacts, the claim that aligned latents preserve generalizable anatomy would be weakened.

Watch

Extended reading notes

Core claim

The paper's central claim is that domain-adapted representation alignment is the primary driver of synthesis fidelity in endoscopic diffusion transformers. Aligning the hidden states of a Scalable Interpolant Transformer (SiT) with patch-level features from a DINOv3-B encoder pretrained on GN-5M lowers FID from 13.55 (no alignment, SD2 VAE, SiT-B/2) to 10.18 with the in-domain teacher, and scaling the backbone to SiT-L/2 on the full dataset reaches 5.32. The same frozen backbone, probed at layer 8 with a linear head, achieves BE AUC 0.786 and POLAR AUC 0.758, beating both EndoViT (0.629/0.632) and Endo-FM (0.764/0.683) and all general-purpose encoders tested; it also keeps higher AUC than both endoscopy-specific models under synthetic imaging corruptions. REVEAL also demonstrates spatially coherent inpainting and outpainting via RePaint-style resampling, which the paper reads as evidence that the aligned latent space respects gastrointestinal morphology rather than just pixel statistics.

Load-bearing premise

The load-bearing premise is that the in-domain teacher, trained on the same corpus the generative student learns, transfers generalizable anatomical structure rather than memorized statistics of GN-5M.

Editorial extensions

If this is right

  • A single generative backbone trained without labels can double as a classification feature extractor, outperforming dedicated endoscopy encoders on polyp and Barrett's esophagus linear probing.
  • Using an in-domain teacher matters more than model size or VAE choice: the DINOv3-B (GN-5M) teacher gives the best FID and POLAR results among all alignment targets tested.
  • Training on the full 4.82M-frame distribution yields the largest single gain in both fidelity and representation quality, and continued training past 200k iterations gives diminishing returns.
  • The same frozen model supports unconditional generation, inpainting, and outpainting without architectural changes, so synthesis and editing tools can be built directly on the released weights.
  • Public release of weights lowers the compute threshold for downstream conditional synthesis, segmentation, and out-of-distribution detection in gastroenterology.

Reading between the lines

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

  • Editorial inference: the same representation-alignment recipe should transfer to other clinical imaging domains that have a large unlabeled archive and a domain-pretrained teacher, since nothing in the method is endoscopy-specific.
  • Editorial inference: because the teacher, the student, the FID reference, and the pretraining distribution all come from GN-5M, the headline fidelity gain may partly reflect familiarity with the training manifold; an external FID set or a blinded clinician study would separate anatomical realism from memorization.
  • Editorial inference: using intermediate diffusion timesteps $t>0$ for feature extraction, which the authors note naturally denoise corrupted inputs, is a promising cheap route to further robustness gains without retraining.
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 / 5 minor

Summary. The paper introduces REVEAL, a latent diffusion transformer (SiT) trained with representation alignment on GN-5M, a 4.82M-frame multicenter endoscopic dataset. The model aligns noisy hidden states with a domain-adapted DINOv3-B teacher using an iREPA-style objective, and the authors ablate VAE backends, teacher encoders, model scale, dataset size, and training duration. They report FID improvements for generation and linear-probing AUC on the POLAR benchmark, a private Barrett's Esophagus dataset (BE), and a corrupted variant (BE-C). The central claims are that REVEAL is the largest generative foundation model for endoscopy to date, that domain-adapted representation alignment improves synthesis fidelity, and that its frozen features are competitive with or exceed endoscopy-specific discriminative foundation models such as EndoViT and Endo-FM.

Significance. If the claims hold, REVEAL would be a valuable contribution: it is the first large-scale generative pretraining for gastrointestinal endoscopy with a domain-adapted teacher, and the paper includes systematic ablations (Table 1) that disentangle latent space, teacher, scale, and data size. The public release of code and weights is a concrete strength, and the use of held-out external datasets (POLAR, BE, BE-C) for the classification claims is a positive feature. However, the synthesis-fidelity claim currently rests on an underspecified FID protocol, the evaluation sets are not shown to be patient-disjoint from the pretraining corpus, and the linear-probing protocol lacks patient-level stratification and variance reporting. These gaps prevent full verification of the paper's central comparative statements.

major comments (4)
  1. [Sec. 4.3, Table 1] The FID protocol is underspecified. The text states only that "FID scores are computed from samples generated using the Euler solver with 50 function evaluations (NFEs)," without reporting the reference image set, its size, its preprocessing, or the number of generated samples used. Since the central fidelity claims (unguided 13.55 vs. aligned 10.18 at 150k iterations; 5.32 at 250k on the full set) depend entirely on these numbers, the authors must specify the reference set, confirm that it is disjoint from the training data, and state the sample count.
  2. [Secs. 4.1-4.3, Tables 2-3] No patient-level or procedure-level deduplication is reported between GN-5M pretraining and the POLAR, BE, and BE-C evaluation sets. GN-5M was collected from eight Dutch hospitals including Amsterdam UMC; the private BE set originates from Amsterdam UMC, and POLAR is an Amsterdam-based public benchmark. Without an explicit exclusion statement, the reported AUC gains over EndoViT and Endo-FM, and the FID improvements if the FID reference is drawn from GN-5M, could reflect memorization of training patients rather than generalization. Please provide a patient-level overlap check or describe how the evaluation sets were kept disjoint from pretraining.
  3. [Sec. 4.3, Tables 2-3] The linear-probing protocol is ambiguous: "We evaluate on the test sets of both benchmarks with five-fold cross-validation" does not state whether the folds are patient-stratified or image-level, whether the probe is trained on the official training split and evaluated on the test split, or how the reported mean is obtained. No per-fold variance or confidence interval is given. Given the small test sets (273 BE images; 588 POLAR images), the reported margins (e.g., REVEAL vs. DINOv2, BE AUC 0.786 vs. 0.772) need variance estimates to support the comparative claims. Please report patient-level stratified cross-validation with per-fold results and confidence intervals.
  4. [Sec. 4.2 and Sec. 5.4] The benefit of the in-domain teacher for synthesis is currently supported only by FID numbers whose reference distribution is unspecified and likely overlaps with GN-5M; the teacher, the student, and the FID reference all derive from the same multicenter corpus. The held-out POLAR and BE benchmarks validate the discriminative features but do not validate the synthesis claim itself. I recommend either computing FID against an external endoscopic reference set (e.g., one of the hospitals not used in training, or POLAR frames) or adding a quantitative external evaluation (e.g., a reader study or a downstream segmentation utility test) to rule out the possibility that the fidelity gains reflect reproduction of training-distribution statistics.
minor comments (5)
  1. [Abstract and Sec. 1] The claim "largest generative foundation model for endoscopy to date" is not substantiated; please specify whether "largest" refers to parameter count, training data size, or both, and compare with prior generative endoscopy models such as Endora and Polyp-DDPM.
  2. [Abstract and throughout] Typographical artifacts should be corrected: "underrealisticimagingcorruptions" in the abstract and "T able 1", "T able 2", and "T able 3" in the body appear to be formatting errors.
  3. [Sec. 5.4] The inpainting and outpainting evidence is purely visual. Since the paper claims "robust structural coherence" in these edits, please either add quantitative metrics (e.g., LPIPS or PSNR on held-out masks) or explicitly describe these results as anecdotal demonstrations.
  4. [Sec. 5.2] The comparison with EndoViT and Endo-FM should acknowledge the input-resolution mismatch: those models were pretrained and evaluated at 224x224, while REVEAL features are extracted at 256x256, which may contribute to the reported differences.
  5. [Sec. 4.2] The two-stage DINOv3-B (GN-5M) training curriculum does not state whether each 115,000-iteration stage uses the full GN-5M dataset or the 250k-image subset; please clarify for reproducibility.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the derivation chain is a standard alignment-plus-denoising objective, and the headline comparisons are evaluated on external benchmarks; the only self-citation is the GN-5M dataset, which is not load-bearing.

full rationale

The paper's derivation chain is explicit: Eq. (1) defines an iREPA/REPA alignment loss, Eq. (2) applies spatial normalization to teacher features, and Eq. (3) combines the alignment loss with a denoising loss. None of these equations define a target quantity in terms of itself, and no fitted parameter is renamed as a prediction. The in-domain teacher DINOv3-B (GN-5M) is an input trained by the authors on GN-5M, and the student is trained on the same corpus; however, the comparison between teachers is empirical and can fail (DINOv2-B (GN-5M) yields worse FID than DINOv2-B, Table 1), so the outcome is not forced by construction. The classification and robustness claims are evaluated with linear probing on the public POLAR and private BE/BE-C test sets, which are external to the generative training objective; moreover, REVEAL underperforms its own teacher (Table 2), showing the discriminative results are not a tautological copy of the teacher. The only notable self-citation is GN-5M [16], whose author list overlaps with the present paper; it is used as a dataset resource rather than as a result invoked to forbid alternatives. The lack of stated patient-level deduplication between GN-5M and the BE/POLAR evaluation sets is a data-independence and leakage risk, not a circularity of the derivation, and is therefore noted as a correctness concern rather than a circular step.

Assumptions & free parameters 5 free parameters · 6 assumptions · 0 invented entities

The ledger is light: REVEAL introduces no new theoretical entities, no new loss functions, and no new architectural primitives. Everything is assembled from SiT, a frozen SD2 VAE, DINOv3, the iREPA alignment module, and the GN-5M corpus. What the paper does not pay for upstream is the data and the alignment premise: (1) the GN-5M corpus must faithfully represent endoscopy; (2) the REPA/iREPA premise that teacher alignment improves generation must hold in this domain; (3) the private BE benchmark must be a fair evaluator; (4) the teacher and student training on the same distribution must not collapse into memorization. The free parameters are inherited hyperparameters, none of which is fitted to the benchmark targets.

free parameters (5)
  • lambda alignment weight = 1
    Eq. 3 balances the denoising and REPA losses. It is fixed at 1 following iREPA [32] and never ablated, so all reported FID gains are conditional on it.
  • logit-normal noise schedule (mu, sigma) = 0.8, 0.8
    Section 4.2 samples logit(t) from N(0.8, 0.8), inherited from prior work [22,25]; it shapes the training-time noise distribution and is not ablated here.
  • Feature extraction layer for classification = 8th SiT layer at t=0
    Section 4.3 fixes the linear-probing features at layer 8 and t=0; this choice affects all downstream accuracy numbers and is not ablated.
  • DINOv3 (GN-5M) teacher curriculum = 2x115k iterations, 16,384 prototypes, mask 0.1-0.5
    Section 4.2 hand-designs a two-stage curriculum for the teacher; the teacher quality is the main driver of the alignment benefit, so these choices are load-bearing.
  • Convolutional projection kernel size = 3
    Section 4.2 uses a 3x3 conv projection in the iREPA module, inherited from iREPA [32] and not ablated in this paper.
assumptions (6)
  • domain assumption GN-5M is a faithful sample of the endoscopic manifold across centers, procedures, and modalities
    Section 4.1: all pretraining, teacher training, and FID evaluations live on GN-5M; any bias in the corpus (capsule vs colonoscopy mix, center effects) propagates into the model.
  • domain assumption Cosine alignment of SiT hidden states to frozen teacher features is a valid proxy for anatomical fidelity
    Eqs. 1-3 adopt the iREPA premise from [32]; the paper provides no independent evidence that alignment scores correlate with clinical morphological correctness.
  • domain assumption The private BE test set with enriched subtle neoplasia is a suitable unbiased evaluation target
    Section 4.1: the BE benchmark comes from the authors' own centers, and the enrichment design is only described in one sentence, leaving unknown difficulty shifts.
  • standard math DINO/iBOT self-supervised objectives remain well-behaved on endoscopic images under the chosen curriculum
    Section 4.2: the teacher training relies on standard DINOv3/DINOv2 machinery (Sinkhorn-Knopp centering, RoPE, iBOT masking) without modification.
  • domain assumption The BE-C corruption protocol reflects realistic clinical imaging artifacts
    Section 4.1: robustness is measured on synthetic perturbations following the same group's protocol [14]; no real-world corrupted clinical corpus is used.
  • domain assumption Frozen SiT hidden states at t=0 encode semantically useful patch descriptors
    Section 3: the feature-extraction mode reads intermediate transformer features at the clean-latent endpoint of the flow; this is an unusual use of a denoiser and is validated only by the benchmark numbers, not by analysis.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Representation-driven Endoscopic Visual Embedding Alignment for Latent Generation." pith.science (2026). https://pith.science/paper/JLR24LBU

@misc{pith2026260807176,
  author       = {Pith},
  title        = {Pith review of: Representation-driven Endoscopic Visual Embedding Alignment for Latent Generation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/JLR24LBU}},
  note         = {Machine review of arXiv:2608.07176}
}
read the original abstract

Developing foundation generative models for endoscopy is limited by the gap between natural and clinical images and the computational cost of training large Diffusion Transformers. Although representation alignment has improved efficiency in general computer vision, its role within the highly specialized endoscopic image space remains unclear. We introduce REVEAL (Representation-driven Endoscopic Visual Embedding Alignment), the largest generative foundation model for endoscopy to date, trained on GastroNet-5M (GN-5M), a multicenter dataset of 5 million endoscopic frames. Instead of depending on out-of-domain priors, REVEAL employs encoders pretrained directly on the endoscopic distribution to align diffusion latents with domain-specific visual features, preserving fine textures and intricate anatomical structures. Beyond image generation, REVEAL also serves as a powerful feature extractor; in multiple benchmarks, it delivers performance that is competitive with, and in several cases exceeds, endoscopic foundation models such as EndoViT and Endo-FM, specifically tuned for classification tasks, while demonstrating strong representation robustness under realistic imaging corruptions. REVEAL produces high-fidelity images and maintains robust structural coherence in latent-space edits such as inpainting and outpainting. This high-capacity backbone lowers the computational threshold for building specialized clinical tools, offering an open, versatile foundation for conditional synthesis, segmentation, and out-of-distribution detection in future intelligent gastroenterology systems.

Figures

Figures reproduced from arXiv: 2608.07176 by the authors.

Figure 1
Figure 1. The REVEAL framework. Training (top) optimizes SiT blocks using joint denoising and feature alignment objectives. Feature extraction (bottom) yields high￾dimensional patch descriptors for downstream analysis. Sampling (right) synthesizes images by traversing the learned reverse SiT trajectory. work has extended this idea in several directions: REPA-E [20] unlocks end-to￾end joint training of the VAE and diffusion mo… view at source ↗
Figure 2
Figure 2. Examples of images from the various datasets used in this study: (a) unlabeled GN-5M images employed for training, illustrating the diversity of the data; (b) polyp images from the POLAR dataset; (c) sample images from the private BE dataset; (d) examples of corrupted BE images. 4 Methodology 4.1 Datasets Pretraining Dataset. We use the GN-5M dataset [16] to pretrain our spe￾cialized vision encoders and to train our… view at source ↗
Figure 3
Figure 3. Qualitative results from REVEAL, generated with a Heun solver using 50 NFEs. The samples demonstrate high-fidelity synthesis of diverse endoscopic scenes [PITH_FULL_IMAGE:figures/full_fig_p013_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

41 extracted references · 21 canonical work pages

  1. [1]

    International Journal of Computer Assisted Radiology and Surgery19(6), 1085–1091 (2024)

    Batić, D., Holm, F., Özsoy, E., Czempiel, T., Navab, N.: Endovit: pretraining vision transformers on a large collection of endoscopic images. International Journal of Computer Assisted Radiology and Surgery19(6), 1085–1091 (2024)

    2024

  2. [2]

    Medical Image Analysis98, 103298 (2024)

    Boers, T.G., Fockens, K.N., van der Putten, J.A., Jaspers, T.J., Kusters, C.H., Jukema, J.B., Jong, M.R., Struyvenberg, M.R., de Groof, J., Bergman, J.J., et al.: Foundation models in gastrointestinal endoscopic ai: Impact of architecture, pre- training approach and data efficiency. Medical Image Analysis98, 103298 (2024)

    work page 2024

  3. [3]

    In: MICCAI Workshop on Deep Generative Models

    Caetano, F., Abdi, L., Viviers, C., Valiuddin, A., van der Sommen, F.: Medsymm- flow: Bridging generative modeling and classification in medical imaging through symmetrical flow matching. In: MICCAI Workshop on Deep Generative Models. pp. 35–45. Springer (2025)

    work page 2025

  4. [4]

    Proceedings of the AAAI Conference on Artificial Intel- ligence40(4), 2498–2506 (Mar 2026).https://doi.org/10.1609/aaai.v40i4

    Caetano, F., Viviers, C., With, P.H.d., Van der Sommen, F.: Symmetrical flow matching: Unified image generation, segmentation, and classification with score- based generative models. Proceedings of the AAAI Conference on Artificial Intel- ligence40(4), 2498–2506 (Mar 2026).https://doi.org/10.1609/aaai.v40i4. 37236,http://dx.doi.org/10.1609/aaai.v40i4.37236

    work page doi:10.1609/aaai.v40i4 2026

  5. [5]

    arXiv preprint arXiv:2002.04185 (2020)

    Chu, C., Minami, K., Fukumizu, K.: Smoothness and stability in gans. arXiv preprint arXiv:2002.04185 (2020)

    arXiv 2020

  6. [6]

    In: 2009 IEEE conference on computer vision and pattern recognition

    Deng, J., Dong, W., Socher, R., Li, L.J., Li, K., Fei-Fei, L.: Imagenet: A large- scale hierarchical image database. In: 2009 IEEE conference on computer vision and pattern recognition. pp. 248–255. Ieee (2009)

    2009

  7. [7]

    IEEE Access12, 25668–25683 (2024)

    Diamantis, D.E., Gatoula, P., Koulaouzidis, A., Iakovidis, D.K.: This intestine does not exist: Multiscale residual variational autoencoder for realistic wireless capsule endoscopy image generation. IEEE Access12, 25668–25683 (2024)

    work page 2024

  8. [8]

    In: 2025 IEEE International Conference on Imaging Systems and Techniques (IST)

    Diamantis, D.E., Iakovidis, D.K.: Multiscale vector-quantized variational autoen- coder for endoscopic image synthesis. In: 2025 IEEE International Conference on Imaging Systems and Techniques (IST). pp. 1–6. IEEE (2025)

    work page 2025

Show all 41 references
  1. [9]

    In: 2024 46th Annual International Confer- ence of the IEEE Engineering in Medicine and Biology Society (EMBC)

    Dorjsembe, Z., Pao, H.K., Xiao, F.: Polyp-ddpm: Diffusion-based semantic polyp synthesis for enhanced segmentation. In: 2024 46th Annual International Confer- ence of the IEEE Engineering in Medicine and Biology Society (EMBC). pp. 1–7. IEEE (2024)

    2024

  2. [10]

    In: Forty-first international conference on machine learning (2024)

    Esser, P., Kulal, S., Blattmann, A., Entezari, R., Müller, J., Saini, H., Levi, Y., Lorenz, D., Sauer, A., Boesel, F., et al.: Scaling rectified flow transformers for high-resolution image synthesis. In: Forty-first international conference on machine learning (2024)

    2024

  3. [11]

    arXiv preprint arXiv:2502.03687 (2025) 16 F

    Favero, G.M., Saremi, P., Kaczmarek, E., Nichyporuk, B., Arbel, T.: Conditional diffusion models are medical image classifiers that provide explainability and un- certainty for free. arXiv preprint arXiv:2502.03687 (2025) 16 F. Caetano et al

    2025 arXiv

  4. [12]

    Computer Vision and Image Understanding240, 103928 (2024)

    Grechka, A., Couairon, G., Cord, M.: Gradpaint: Gradient-guided inpainting with diffusion models. Computer Vision and Image Understanding240, 103928 (2024)

    2024

  5. [13]

    Endoscopy55(08), 756–765 (2023)

    Houwen, B.B., Hazewinkel, Y., Giotis, I., Vleugels, J.L., Mostafavi, N.S., van Put- ten, P., Fockens, P., Dekker, E., Group, P.S., et al.: Computer-aided diagnosis for optical diagnosis of diminutive colorectal polyps including sessile serrated lesions: a real-time comparison ...

    2023

  6. [14]

    Medical Image Analysis94, 103157 (2024)

    Jaspers, T.J., Boers, T.G., Kusters, C.H., Jong, M.R., Jukema, J.B., De Groof, A.J., Bergman, J.J., de With, P.H., van der Sommen, F.: Robustness evaluation of deep neural networks for endoscopic image analysis: Insights and strategies. Medical Image Analysis94, 103157 (2024)

    2024

  7. [15]

    In: MICCAI Workshop on Data Engineering in Medical Imaging

    Jaspers, T.J., Caetano, F., Claessens, C.H., Kusters, C.H., Middeljans, H., Jong, M.R., van Eijck van Heslinga, R.A., Slooter, F., de Groof, A.J., Bergman, J.J., et al.: Robust early detection of barrett’s neoplasia: Addressing low-prevalence challenges with generative modelin...

    2025

  8. [16]

    Gastroenterology (2025)

    Jong, M.R., Boers, T.G., Fockens, K.N., Jukema, J.B., Kusters, C.H., Jaspers, T.J., van Heslinga, R.v.E., Slooter, F.C., Struyvenberg, M.R., Bisschops, R., et al.: Gastronet-5m: A multicenter dataset for developing foundation models in gastroin- testinal endoscopy. Gastroenter...

    2025

  9. [17]

    Kaleta, J., Dall’Alba, D., Płotka, S., Korzeniowski, P.: Minimal data requirement forrealisticendoscopicimagegenerationwithstablediffusion.Internationaljournal of computer assisted radiology and surgery19(3), 531–539 (2024)

    2024

  10. [18]

    Labs, B.F.: Announcing black forest labs.https://bfl.ai/blog/24-08-01-bfl (2024), accessed: 2026-02-06

    2024

  11. [19]

    Labs, B.F.: Flux.2: Frontier visual intelligence.https://bfl.ai/blog/flux- 2 (2025), accessed: 2026-02-06

    2025

  12. [20]

    In: Proceedings of the IEEE/CVF International Conference on Computer Vision

    Leng, X., Singh, J., Hou, Y., Xing, Z., Xie, S., Zheng, L.: Repa-e: Unlocking vae for end-to-end tuning of latent diffusion transformers. In: Proceedings of the IEEE/CVF International Conference on Computer Vision. pp. 18262–18272 (2025)

    2025

  13. [21]

    In: International conference on medical image computing and computer-assisted intervention

    Li, C., Liu, H., Liu, Y., Feng, B.Y., Li, W., Liu, X., Chen, Z., Shao, J., Yuan, Y.: Endora: Video generation models as endoscopy simulators. In: International conference on medical image computing and computer-assisted intervention. pp. 230–240. Springer (2024)

    2024

  14. [22]

    arXiv preprint arXiv:2511.13720 (2025)

    Li, T., He, K.: Back to basics: Let denoising generative models denoise. arXiv preprint arXiv:2511.13720 (2025)

    2025 arXiv

  15. [23]

    In: 2025 IEEE International Conference on Robotics and Automation (ICRA)

    Liu,S.,Chen,Z.,Yang,Q.,Yu,W.,Dong,D.,Hu,J.,Yuan,Y.:Polyp-gen:Realistic and diverse polyp image generation for endoscopic dataset expansion. In: 2025 IEEE International Conference on Robotics and Automation (ICRA). pp. 15776– 15782. IEEE (2025)

    2025

  16. [24]

    In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

    Lugmayr, A., Danelljan, M., Romero, A., Yu, F., Timofte, R., Van Gool, L.: Re- paint: Inpainting using denoising diffusion probabilistic models. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 11461– 11471 (2022)

    2022

  17. [25]

    In: European Conference on Computer Vision

    Ma, N., Goldstein, M., Albergo, M.S., Boffi, N.M., Vanden-Eijnden, E., Xie, S.: Sit: Exploring flow and diffusion-based generative models with scalable interpolant transformers. In: European Conference on Computer Vision. pp. 23–40. Springer (2024)

    2024

  18. [26]

    In: Proceedings of the 4th ACM Workshop on Intelligent Cross-Data Analysis and Retrieval

    Macháček, R., Mozaffari, L., Sepasdar, Z., Parasa, S., Halvorsen, P., Riegler, M.A., Thambawita, V.: Mask-conditioned latent diffusion for generating gastrointestinal REVEAL 17 polyp images. In: Proceedings of the 4th ACM Workshop on Intelligent Cross-Data Analysis and Retriev...

    2023

  19. [27]

    arXiv preprint arXiv:2304.07193 (2023)

    Oquab, M., Darcet, T., Moutakanni, T., Vo, H., Szafraniec, M., Khalidov, V., Fernandez, P., Haziza, D., Massa, F., El-Nouby, A., et al.: Dinov2: Learning robust visual features without supervision. arXiv preprint arXiv:2304.07193 (2023)

    2023 arXiv

  20. [28]

    arXiv preprint arXiv:2408.00714 (2024)

    Ravi, N., Gabeur, V., Hu, Y.T., Hu, R., Ryali, C., Ma, T., Khedr, H., Rädle, R., Rolland, C., Gustafson, L., et al.: Sam 2: Segment anything in images and videos. arXiv preprint arXiv:2408.00714 (2024)

    2024 arXiv

  21. [29]

    In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition

    Rombach, R., Blattmann, A., Lorenz, D., Esser, P., Ommer, B.: High-resolution image synthesis with latent diffusion models. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. pp. 10684–10695 (2022)

    2022

  22. [30]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition

    Sharma, V., Kumar, A., Jha, D., Bhuyan, M.K., Das, P.K., Bagci, U.: Con- trolpolypnet: towards controlled colon polyp synthesis for improved polyp segmen- tation. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 2325–2334 (2024)

    2024

  23. [31]

    arXiv preprint arXiv:2508.10104 (2025)

    Siméoni, O., Vo, H.V., Seitzer, M., Baldassarre, F., Oquab, M., Jose, C., Khali- dov, V., Szafraniec, M., Yi, S., Ramamonjisoa, M., et al.: Dinov3. arXiv preprint arXiv:2508.10104 (2025)

    2025 arXiv

  24. [32]

    Singh, J., Leng, X., Wu, Z., Zheng, L., Zhang, R., Shechtman, E., Xie, S.: What matters for representation alignment: Global information or spatial structure? arXiv preprint arXiv:2512.10794 (2025)

    2025

  25. [33]

    In: International Conference on Medical Image Computing and Computer-Assisted Intervention

    Tian, Q., Liao, H., Huang, X., Yang, B., Lei, D., Ourselin, S., Liu, H.: Endomamba: an efficient foundation model for endoscopic videos via hierarchical pre-training. In: International Conference on Medical Image Computing and Computer-Assisted Intervention. pp. 224–234. Sprin...

    2025

  26. [34]

    arXiv preprint arXiv:1607.08022 (2016)

    Ulyanov, D., Vedaldi, A., Lempitsky, V.: Instance normalization: The missing in- gredient for fast stylization. arXiv preprint arXiv:1607.08022 (2016)

    2016 arXiv

  27. [35]

    Wang, Z., Liu, C., Zhang, S., Dou, Q.: Foundation model for endoscopy video anal- ysisvialarge-scaleself-supervisedpre-train.In:Internationalconferenceonmedical image computing and computer-assisted intervention. pp. 101–111. Springer (2023)

    2023

  28. [36]

    IEEE Journal of Biomedical and Health Informatics29(5), 3526–3536 (2025)

    Wang, Z., Liu, C., Zhu, L., Wang, T., Zhang, S., Dou, Q.: Improving foundation model for endoscopy video analysis via representation learning on long sequences. IEEE Journal of Biomedical and Health Informatics29(5), 3526–3536 (2025)

    2025

  29. [37]

    generation: Taming optimization dilemma in latent diffusion models

    Yao, J., Yang, B., Wang, X.: Reconstruction vs. generation: Taming optimization dilemma in latent diffusion models. In: Proceedings of the Computer Vision and Pattern Recognition Conference. pp. 15703–15712 (2025)

    2025

  30. [38]

    In: The Thirteenth International Conference on Learning Representations (2025)

    Yu, S., Kwak, S., Jang, H., Jeong, J., Huang, J., Shin, J., Xie, S.: Representation alignment for generation: Training diffusion transformers is easier than you think. In: The Thirteenth International Conference on Learning Representations (2025)

    2025

  31. [39]

    In: The Thirty-ninth Annual Conference on Neural Information Processing Systems (2025)

    Zhang, J., Shen, Y., Chen, G., Song, L., Xing, E.P.: Dimensional collapse in vq- vaes: Evidence and remedies. In: The Thirty-ninth Annual Conference on Neural Information Processing Systems (2025)

    2025

  32. [40]

    arXiv preprint arXiv:2508.05772 (2025)

    Zhao, C., Guo, P., Yang, D., Tang, Y., He, Y., Simon, B., Belue, M., Harmon, S., Turkbey, B., Xu, D.: Maisi-v2: Accelerated 3d high-resolution medical image synthesis with rectified flow and region-specific contrastive loss. arXiv preprint arXiv:2508.05772 (2025)

    2025

  33. [41]

    arXiv preprint arXiv:2502.03491 (2025)

    Zheng, C., Lan, Y., Wang, Y.: Lanpaint: Training-free diffusion inpaint- ing with asymptotically exact and fast conditional sampling. arXiv preprint arXiv:2502.03491 (2025)

    2025

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.