Pith. sign in

REVIEW 3 major objections 4 minor 39 references

An Ordinary Differential Equation Sampler with Stochastic Start for Diffusion Bridge Models

T0 review · 3 major / 4 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 The paper claims that pure ODE samplers in diffusion bridge models fail because the probability flow ODE is singular at the start, and fixes it with a Gaussian stochastic start followed by Heun's second-order solver.

desk verdict A clever, well-written sampler for diffusion bridges with a genuine theoretical gap in the stochastic-start justification; worth peer review with requests for ablations. read the letter →

arxiv 2412.19992 v1 pith:ERSYLQW4 submitted 2024-12-28 cs.CV cs.AI

Yuang Wang , Pengfei Jin , Li Zhang , Quanzheng Li , Zhiqiang Chen , Dufan Wu This is my paper · ORCID
classification cs.CVcs.AI
keywords diffusionbridgemodelprobabilityflowODEsamplerposteriorsamplingHeun'smethodimagerestorationtranslationstochasticstart
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

Diffusion bridge models generate images by starting from a corrupted input rather than pure noise, but they have typically required slow stochastic (SDE) samplers. This paper argues that deterministic ODE samplers underperform on these models because the probability flow ODE is singular at the very first step, where the process is pinned to the corrupted image. The proposed sampler, ODES3, sidesteps the singularity by drawing the first intermediate state from a Gaussian posterior centered near the predicted clean image, then integrates the well-behaved ODE with Heun's second-order solver. On super-resolution, JPEG restoration, and two image-translation tasks, this training-free recipe matches or beats prior samplers while using 28–38 neural function evaluations instead of 100 or more.

What carries the argument

The paper's central object is the probability flow ODE (PF-ODE) of a diffusion bridge model, whose drift term combines a learned score with an explicit h-transform term. Theorems 1 and 2 characterize its boundary behavior: the drift is well-defined at the terminal time for the reverse SDE but singular for the PF-ODE. To bypass the singularity, the sampler uses a stochastic start defined by the Gaussian posterior q_{τ|0,y}(X_τ | \hat{X}$_0^{{(T)}}$, y), where \hat{X}$_0^{{(T)}}$ = D_θ(y, y, T) is the trained data predictor's estimate of the clean image. This single draw replaces the first Euler step; then Heun's second-order method, an explicit two-stage Runge–Kutta scheme, integrates the PF-ODE.

What would settle it

Compute the empirical covariance of X_τ from many independent draws of the reverse SDE at time τ and compare it with the covariance $c_τ^{2}$ I used in the stochastic start; a large excess variance would indicate the start is under-dispersed and the deterministic trajectory is biased.

Watch

Extended reading notes

Core claim

The central claim is that the limited performance of pure ODE samplers in diffusion bridge models is caused by the singular behavior of the PF-ODE at the start of the generative process. The paper shows that the non-linear drift in the PF-ODE has no finite limit as time approaches the terminal point, while the reverse SDE's drift is well-defined. To jump over the singularity, it introduces posterior sampling at the first reverse step: X_τ is drawn from q_{τ|0,y}(X_τ | \hat{X}$_0^{{(T)}}$, y), a Gaussian whose mean uses the data predictor's estimate of the clean image and whose covariance is the analytic $c_τ^{2}$ I. The paper proves this start is closer in KL divergence to the exact transition distribution than a single Euler–Maruyama step. After the jump, Heun's second-order solver integrates the PF-ODE from τ to 0, yielding the reported gains in FID and speed.

Load-bearing premise

The load-bearing premise is that sampling X_τ from a single Gaussian centered at the predicted clean image, with covariance $c_τ^{2}$ I, is a faithful stand-in for the true marginal posterior, whose extra spread from the unknown clean image is ignored.

Editorial extensions

If this is right

  • ODES3 matches or beats the original samplers of I2SB and DDBM on all four tasks with far fewer neural function evaluations.
  • The method requires no additional training and works with already-pretrained diffusion bridge models.
  • It reduces FID by 9–18% on restoration tasks and by 11–70% on translation tasks compared to prior samplers.
  • Inference time per image drops by roughly 2.7× on restoration tasks and 3.6–4.2× on translation tasks.

Reading between the lines

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

  • The stochastic start uses a covariance that ignores the spread of clean images around the predicted mean; a natural testable extension is to inflate the start covariance or learn it, which might help tasks with highly ambiguous degradations.
  • The choice of the jump time τ is taken from the pretrained model's schedule; tuning τ directly could trade off singularity avoidance against the distance the ODE must integrate.
  • The same stochastic-start idea could be applied to other conditional generation settings where an ODE endpoint is singular, such as inpainting or colorization.
  • Since Theorem 3 optimizes a conditional KL, the start is optimal only within the Gaussian family; a mixture-based start might do better when the posterior is multimodal.
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

3 major / 4 minor

Summary. The manuscript proposes ODES3, a training-free ODE sampler for diffusion bridge models. It observes that the PF-ODE for diffusion bridges has a singular drift at the terminal time T, while the reverse SDE is well-defined. To initialize the ODE trajectory, the method samples X_τ from a Gaussian posterior q_post defined by plugging the data predictor's estimate of the clean image into the transition kernel (Eq. 14), then integrates the PF-ODE from τ to 0 with Heun's second-order solver. Experiments on super-resolution, JPEG restoration, and two image-to-image translation tasks show improved FID with 28–38 NFEs compared to SDE or hybrid samplers using 100+ NFEs.

Significance. If the theoretical justification were complete, this would be a valuable contribution: it is a simple, training-free acceleration applicable to pretrained bridge models, with clean singularity theorems (Thms 1–2) and a correct conditional-KL computation (Thm 3, App. A.2). The empirical results are strong and the method is fully reproducible given the released pretrained models. However, the central theoretical claim—that the stochastic start q_post is the appropriate marginal for initializing the PF-ODE—has a gap, because Theorem 3 optimizes a conditional KL rather than the marginal KL. This needs to be addressed before the method's theoretical basis is solid.

major comments (3)
  1. [Section III-B, Eq. (14), Appendix A.2] The distribution q_post(X_τ|y) = q_{τ|0,y}(X_τ | \hat{X}_0^{(T)}, y) has covariance c_τ^2 I, but the true marginal start needed by the PF-ODE (9) is q_{τ|y}(X_τ|y) = E_{X_0|y}[q_{τ|0,y}(X_τ|X_0,y)], whose covariance is c_τ^2 I + b_τ^2 Cov(X_0|y). Theorem 3 solves min_q E_{X_0}[KL(q || q_{τ|0,y}(·|X_0,y))] under a Gaussian constraint; this objective does not include the between-X_0 covariance term, so the optimal covariance is c_τ^2 I. Because the subsequent ODE is deterministic, under-dispersion of the initial condition is never corrected. The paper provides no bound on b_τ^2 Cov(X_0|y) at the chosen τ (e.g., by choosing τ close enough to T that b_τ is small) and no experiment measuring the sensitivity to this term. This is a load-bearing gap, as the stochastic start is the paper's main contribution.
  2. [Theorem 3 and Eq. (15)] The comparison to q_EM in Theorem 3 does not resolve the marginal issue. q_EM is also not the marginal q_{τ|y}; inequality (15) compares expected conditional KLs, and both distributions could be poor approximations to the marginal. A correct justification would either optimize KL(q || q_{τ|y}) directly, or show that b_τ^2 Cov(X_0|y) is negligible for the chosen schedule and tasks.
  3. [Section IV, Algorithm 1] No ablation isolates the effect of the stochastic start. The reported gains could in principle come from the use of Heun's solver or from the specific choice of τ rather than from posterior sampling. I recommend an ablation comparing (i) pure Euler/Heun ODE from a collapsed start (e.g., mean a_τ y + b_τ \hat{X}_0^{(T)} without noise), (ii) q_EM stochastic start, and (iii) the proposed q_post start, all with the same solver and time schedule. This would directly support the causal claim in the abstract and Section I.
minor comments (4)
  1. [Appendix A.2] The phrase 'To proof Theorem 3' should be 'To prove Theorem 3'; similarly, 'inequity (15)' should be 'inequality (15)'.
  2. [Abstract and Introduction] The name 'Frechet' should be spelled 'Fréchet'.
  3. [Algorithm 1] The comment 'SDE sampling at T' on line 4 is misleading; the step is posterior sampling, not SDE discretization. Consider renaming it to 'Posterior sampling at T'.
  4. [Table III] The footnote that baseline results are taken from DDBM and DBIM without exact NFE should be stated more prominently in the main text, since a reader might otherwise treat those rows as directly comparable in NFE.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity; the derivation is self-contained and the only self-citation is non-load-bearing.

full rationale

The paper's central derivation is self-contained rather than circular. Theorems 1 and 2 compute the score and the non-linear drift terms directly from the transition kernel qt|0,y in Eq. (6), and they identify the PF-ODE singularity analytically. Theorem 3 solves the Gaussian-constrained KL minimization in Eqs. (21)-(22) and obtains the mean a_tau y + b_tau X-hat_0^(T) and covariance c_tau^2 I; this is a mathematical derivation, not an evaluation of a quantity fitted to the downstream FID results. The stochastic start qpost uses the pretrained data predictor D_theta* to form X-hat_0^(T), which is an input from earlier external models (I2SB, DDBM) rather than a parameter fitted to the benchmark metrics, so it does not fit the 'fitted input called prediction' pattern. The only self-citation is [24] (I3SB), used in a related-work sentence about non-Markovian samplers; it is not load-bearing, and the paper's claims are tested against external baselines and pretrained checkpoints. The potential weakness that Theorem 3 compares qpost with the conditional q_{tau|0,y}(X_tau|X0,y) rather than the marginal q_{tau|y}(X_tau|y), omitting b_tau^2 Cov(X0|y), is a correctness or robustness question about the initialization, not a circularity: even if the theorem is incomplete, the proposed distribution is not defined in terms of the result it is used to predict. No equation reduces to its own input, and no prediction is forced by a fitted constant or by a self-citation chain.

Assumptions & free parameters 2 free parameters · 3 assumptions · 0 invented entities

The sampler's central claims rest on two assumptions: the pretrained data predictor remains accurate at the singular endpoint t=T, and the KL-optimal Gaussian start is sufficient despite its under-dispersed covariance. The number of steps N and the time schedule are user-chosen hyperparameters.

free parameters (2)
  • N (number of generative steps) = 20 for restoration, 15 for translation
    Set by the authors to achieve the reported NFE; not derived from theory and not swept.
  • Time schedule including τ = Inherited from I2SB and DDBM, exact discretization not reported
    The algorithm requires a schedule 0=t0<...<tN=T; the paper says it follows the pretrained models' schedules but does not specify the exact time points, making the first transition time τ a hidden choice.
assumptions (3)
  • domain assumption The pretrained data predictor Dθ provides an accurate estimate of the conditional mean E[X0|Xt,y] at all times, including t=T where it was not trained.
    The stochastic start uses Dθ(y,y,T) as \hat X0^{(T)}; the model was trained on noisy samples with t<T, so extrapolation to T is assumed valid. Section III-B, Algorithm 1.
  • standard math The score function ∇ log q_{t|y} is smooth and interchanging derivative and integral in the Appendix is valid for t<T.
    Standard regularity conditions for diffusion score matching; not stated explicitly but required for Eq. (16) and the proof of Theorem 1.
  • ad hoc to paper The KL optimality in Theorem 3 is the right criterion for sampling quality.
    The theorem establishes q_post as the optimal Gaussian with respect to a conditional KL, but the link to the true marginal distribution and downstream image quality is assumed rather than proven.

how reviews work

0 comments
Cite this review

Pith. "Pith review of An Ordinary Differential Equation Sampler with Stochastic Start for Diffusion Bridge Models." pith.science (2026). https://pith.science/paper/ERSYLQW4

@misc{pith2026241219992,
  author       = {Pith},
  title        = {Pith review of: An Ordinary Differential Equation Sampler with Stochastic Start for Diffusion Bridge Models},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ERSYLQW4}},
  note         = {Machine review of arXiv:2412.19992}
}
read the original abstract

Diffusion bridge models have demonstrated promising performance in conditional image generation tasks, such as image restoration and translation, by initializing the generative process from corrupted images instead of pure Gaussian noise. However, existing diffusion bridge models often rely on Stochastic Differential Equation (SDE) samplers, which result in slower inference speed compared to diffusion models that employ high-order Ordinary Differential Equation (ODE) solvers for acceleration. To mitigate this gap, we propose a high-order ODE sampler with a stochastic start for diffusion bridge models. To overcome the singular behavior of the probability flow ODE (PF-ODE) at the beginning of the reverse process, a posterior sampling approach was introduced at the first reverse step. The sampling was designed to ensure a smooth transition from corrupted images to the generative trajectory while reducing discretization errors. Following this stochastic start, Heun's second-order solver is applied to solve the PF-ODE, achieving high perceptual quality with significantly reduced neural function evaluations (NFEs). Our method is fully compatible with pretrained diffusion bridge models and requires no additional training. Extensive experiments on image restoration and translation tasks, including super-resolution, JPEG restoration, Edges-to-Handbags, and DIODE-Outdoor, demonstrated that our sampler outperforms state-of-the-art methods in both visual quality and Frechet Inception Distance (FID).

Figures

Figures reproduced from arXiv: 2412.19992 by the authors.

Figure 1
Figure 1. Overview of the proposed ODE sampler with a stochastic start for diffusion bridge models. The forward SDE maps the conditional data distribution [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Visualization results of tested methods for the sr4x-bicubic task. The details within the blue boxes are zoomed in for enhanced visual clarity. The [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Visualization results of tested methods for the JPEG-10 task. The details within the blue and yellow boxes are zoomed in for enhanced visual clarity. [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Visualization results for the Edges→Handbags (64×64) task [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: Visualization results for the DIODE-Outdoor (256 [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

39 extracted references · 7 canonical work pages

  1. [1]

    Score-based generative modeling through stochastic differ- ential equations,

    Y . Song, J. Sohl-Dickstein, D. P. Kingma, A. Kumar, S. Ermon, and B. Poole, “Score-based generative modeling through stochastic differ- ential equations,” arXiv preprint arXiv:2011.13456 , 2020

    arXiv 2011

  2. [2]

    Diffusion models beat gans on image synthesis,

    P. Dhariwal and A. Nichol, “Diffusion models beat gans on image synthesis,” Advances in neural information processing systems , vol. 34, pp. 8780–8794, 2021

    2021

  3. [3]

    Elucidating the design space of diffusion-based generative models,

    T. Karras, M. Aittala, T. Aila, and S. Laine, “Elucidating the design space of diffusion-based generative models,” Advances in neural infor- mation processing systems , vol. 35, pp. 26 565–26 577, 2022

    2022

  4. [4]

    Analyzing and improving the training dynamics of diffusion models,

    T. Karras, M. Aittala, J. Lehtinen, J. Hellsten, T. Aila, and S. Laine, “Analyzing and improving the training dynamics of diffusion models,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024, pp. 24 174–24 184

    work page 2024

  5. [5]

    Real-esrgan: Training real- world blind super-resolution with pure synthetic data,

    X. Wang, L. Xie, C. Dong, and Y . Shan, “Real-esrgan: Training real- world blind super-resolution with pure synthetic data,” in Proceedings of the IEEE/CVF international conference on computer vision , 2021, pp. 1905–1914

    2021

  6. [6]

    Image-to-image translation with conditional adversarial networks,

    P. Isola, J.-Y . Zhu, T. Zhou, and A. A. Efros, “Image-to-image translation with conditional adversarial networks,” in Proceedings of the IEEE conference on computer vision and pattern recognition, 2017, pp. 1125– 1134

    2017

  7. [7]

    I 2sb: Image-to-image schr ¨odinger bridge,

    G.-H. Liu, A. Vahdat, D.-A. Huang, E. A. Theodorou, W. Nie, and A. Anandkumar, “I 2sb: Image-to-image schr ¨odinger bridge,” arXiv preprint arXiv:2302.05872, 2023

    arXiv 2023

  8. [8]

    Denoising diffusion bridge models,

    L. Zhou, A. Lou, S. Khanna, and S. Ermon, “Denoising diffusion bridge models,” arXiv preprint arXiv:2309.16948 , 2023

    arXiv 2023

Show all 39 references
  1. [9]

    Inversion by direct iteration: An al- ternative to denoising diffusion for image restoration,

    M. Delbracio and P. Milanfar, “Inversion by direct iteration: An al- ternative to denoising diffusion for image restoration,” arXiv preprint arXiv:2303.11435, 2023

    2023 arXiv

  2. [10]

    Direct diffusion bridge using data consistency for inverse problems,

    H. Chung, J. Kim, and J. C. Ye, “Direct diffusion bridge using data consistency for inverse problems,” Advances in Neural Information Processing Systems, vol. 36, 2024

    2024

  3. [11]

    Dpm-solver: A fast ode solver for diffusion probabilistic model sampling in around 10 steps,

    C. Lu, Y . Zhou, F. Bao, J. Chen, C. Li, and J. Zhu, “Dpm-solver: A fast ode solver for diffusion probabilistic model sampling in around 10 steps,” Advances in Neural Information Processing Systems , vol. 35, pp. 5775–5787, 2022

    2022

  4. [12]

    Pseudo numerical methods for diffusion models on manifolds,

    L. Liu, Y . Ren, Z. Lin, and Z. Zhao, “Pseudo numerical methods for diffusion models on manifolds,” arXiv preprint arXiv:2202.09778, 2022

    2022 arXiv

  5. [13]

    U. M. Ascher and L. R. Petzold, Computer methods for ordinary differential equations and differential-algebraic equations. SIAM, 1998

    1998

  6. [14]

    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

  7. [15]

    Denoising diffusion implicit models,

    J. Song, C. Meng, and S. Ermon, “Denoising diffusion implicit models,” arXiv preprint arXiv:2010.02502 , 2020

    2010 arXiv

  8. [16]

    Denoising diffusion probabilistic models,

    J. Ho, A. Jain, and P. Abbeel, “Denoising diffusion probabilistic models,” Advances in neural information processing systems , vol. 33, pp. 6840– 6851, 2020

    2020

  9. [17]

    Progressive distillation for fast sampling of diffusion models,

    T. Salimans and J. Ho, “Progressive distillation for fast sampling of diffusion models,” arXiv preprint arXiv:2202.00512 , 2022

    2022 arXiv

  10. [18]

    Consistency models,

    Y . Song, P. Dhariwal, M. Chen, and I. Sutskever, “Consistency models,”

  11. [19]

    Tract: Denoising diffusion models with transitive closure time-distillation,

    D. Berthelot, A. Autef, J. Lin, D. A. Yap, S. Zhai, S. Hu, D. Zheng, W. Talbott, and E. Gu, “Tract: Denoising diffusion models with transitive closure time-distillation,” arXiv preprint arXiv:2303.04248 , 2023

    2023 arXiv

  12. [20]

    Image super-resolution via iterative refinement,

    C. Saharia, J. Ho, W. Chan, T. Salimans, D. J. Fleet, and M. Norouzi, “Image super-resolution via iterative refinement,” IEEE Transactions on Pattern Analysis and Machine Intelligence , vol. 45, no. 4, pp. 4713– 4726, 2022

    2022

  13. [21]

    Palette: Image-to-image diffusion models,

    C. Saharia, W. Chan, H. Chang, C. Lee, J. Ho, T. Salimans, D. Fleet, and M. Norouzi, “Palette: Image-to-image diffusion models,” in ACM SIGGRAPH 2022 conference proceedings , 2022, pp. 1–10

    2022

  14. [22]

    Consistency diffusion bridge models,

    G. He, K. Zheng, J. Chen, F. Bao, and J. Zhu, “Consistency diffusion bridge models,” arXiv preprint arXiv:2410.22637 , 2024

    2024 arXiv

  15. [23]

    Image restoration with mean-reverting stochastic differential equations,

    Z. Luo, F. K. Gustafsson, Z. Zhao, J. Sj ¨olund, and T. B. Sch ¨on, “Image restoration with mean-reverting stochastic differential equations,” arXiv preprint arXiv:2301.11699, 2023

    2023 arXiv

  16. [24]

    Implicit image-to-image schrodinger bridge for ct super-resolution and denoising,

    Y . Wang, S. Yoon, P. Jin, M. Tivnan, Z. Chen, R. Hu, L. Zhang, Z. Chen, Q. Li, and D. Wu, “Implicit image-to-image schrodinger bridge for ct super-resolution and denoising,” arXiv preprint arXiv:2403.06069, 2024

    2024 arXiv

  17. [25]

    Diffusion bridge implicit models,

    K. Zheng, G. He, J. Chen, F. Bao, and J. Zhu, “Diffusion bridge implicit models,” arXiv preprint arXiv:2405.15885 , 2024

    2024 arXiv

  18. [26]

    Imagenet: A large-scale hierarchical image database,

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

    2009

  19. [27]

    Diode: A dense indoor and outdoor depth dataset,

    I. Vasiljevic, N. Kolkin, S. Zhang, R. Luo, H. Wang, F. Z. Dai, A. F. Daniele, M. Mostajabi, S. Basart, M. R. Walter et al., “Diode: A dense indoor and outdoor depth dataset,” arXiv preprint arXiv:1908.00463 , 2019

    1908 arXiv

  20. [28]

    Zero-shot image restoration using denoising diffusion null-space model,

    Y . Wang, J. Yu, and J. Zhang, “Zero-shot image restoration using denoising diffusion null-space model,” arXiv preprint arXiv:2212.00490, 2022

    2022 arXiv

  21. [29]

    Denoising diffusion restoration models,

    B. Kawar, M. Elad, S. Ermon, and J. Song, “Denoising diffusion restoration models,” Advances in Neural Information Processing Sys- tems, vol. 35, pp. 23 593–23 606, 2022

    2022

  22. [30]

    Pseudoinverse-guided diffusion models for inverse problems,

    J. Song, A. Vahdat, M. Mardani, and J. Kautz, “Pseudoinverse-guided diffusion models for inverse problems,” in International Conference on Learning Representations, 2023

    2023

  23. [31]

    Diffusion posterior sampling for general noisy inverse problems,

    H. Chung, J. Kim, M. T. Mccann, M. L. Klasky, and J. C. Ye, “Diffusion posterior sampling for general noisy inverse problems,” arXiv preprint arXiv:2209.14687, 2022

    2022 arXiv

  24. [32]

    Jpeg artifact cor- rection using denoising diffusion restoration models,

    B. Kawar, J. Song, S. Ermon, and M. Elad, “Jpeg artifact cor- rection using denoising diffusion restoration models,” arXiv preprint arXiv:2209.11888, 2022

    2022 arXiv

  25. [33]

    Dual diffusion implicit bridges for image-to-image translation,

    X. Su, J. Song, C. Meng, and S. Ermon, “Dual diffusion implicit bridges for image-to-image translation,” arXiv preprint arXiv:2203.08382, 2022

    2022 arXiv

  26. [34]

    Sdedit: Guided image synthesis and editing with stochastic differential equations,

    C. Meng, Y . He, Y . Song, J. Song, J. Wu, J.-Y . Zhu, and S. Ermon, “Sdedit: Guided image synthesis and editing with stochastic differential equations,” arXiv preprint arXiv:2108.01073 , 2021

    2021 arXiv

  27. [35]

    Flow straight and fast: Learning to generate and transfer data with rectified flow,

    X. Liu, C. Gong, and Q. Liu, “Flow straight and fast: Learning to generate and transfer data with rectified flow,” arXiv preprint arXiv:2209.03003, 2022

    2022 arXiv

  28. [36]

    Deep residual learning for image recognition,

    K. He, X. Zhang, S. Ren, and J. Sun, “Deep residual learning for image recognition,” in Proceedings of the IEEE conference on computer vision and pattern recognition , 2016, pp. 770–778

    2016

  29. [37]

    A note on the inception score,

    S. Barratt and R. Sharma, “A note on the inception score,” arXiv preprint arXiv:1801.01973, 2018

    2018 arXiv

  30. [38]

    The unreasonable effectiveness of deep features as a perceptual metric,

    R. Zhang, P. Isola, A. A. Efros, E. Shechtman, and O. Wang, “The unreasonable effectiveness of deep features as a perceptual metric,” in CVPR, 2018

    2018

  31. [2023]

    Available: https://arxiv.org/abs/2303.01469

    [Online]. Available: https://arxiv.org/abs/2303.01469

    arXiv

Pith tools

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