Pith. sign in

REVIEW 3 major objections 5 minor 63 references

INDIGO+: A Unified INN-Guided Probabilistic Diffusion Algorithm for Blind and Non-Blind Image Restoration

T0 review · 3 major / 5 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read The paper proposes INDIGO and BlindINDIGO, which train a lightweight invertible neural network to simulate any degradation and then use its inverse, with the measurement substituted into the coarse channel, to guide pretrained diffusion…

desk verdict A plausible and well-tested extension of the authors' MMSP 2023 INDIGO to blind restoration, with the main open risk being the unproven coarse/detail disentanglement that the whole guidance step relies on. read the letter →

arxiv 2501.14014 v1 pith:IYDMQVLY submitted 2025-01-23 cs.CV eess.IV

Di You , Pier Luigi Dragotti This is my paper · ORCID
classification cs.CVeess.IV
keywords imagerestorationblinddiffusionmodelsinvertibleneuralnetworksposteriorsamplingdegradationestimationsuper-resolutionDDIMacceleration
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

Most diffusion-based restoration methods need a closed-form expression of the degradation in the non-blind setting, or a fixed family of predefined degradations in the blind setting. This paper proposes INDIGO and BlindINDIGO to remove both restrictions: a small invertible neural network (INN) is trained, purely from paired images, so that its forward pass simulates whatever degradation produced the measurement. At sampling time the network's coarse channel is replaced by the actual measurement and the inverse transform yields an intermediate image that guides the pretrained diffusion model through a gradient step. The authors' central claim is that this yields competitive restorations on synthetic and real-world low-quality images, including non-linear degradations such as downsampling combined with JPEG compression, without retraining the diffusion model.

What carries the argument

The load-bearing object is a lifting-inspired invertible neural network, i.e., a wavelet-lifting construction in which the predict and update steps are neural networks, so that the mapping $[c,d] = f_\phi(x)$ is exactly invertible. Its job is to separate each image into a coarse part that matches the degraded measurement and a detail part that holds the lost information. At inference the coarse part is replaced by the measurement $y$, the inverse $f_\phi^{-1}(y, d_t)$ produces a data-consistent intermediate image, and the distance to the diffusion estimate $x_{0,t}$ is back-propagated into the current noisy sample $x_t$ via a gradient step. In the blind version the same blocks are conditioned on a degradation embedding $\gamma_{\rm deg}$ and finetuned during sampling.

What would settle it

Train the INN on a single synthetic degradation such as 4x bicubic downsampling with Eq. (12). For a set of test images, compute $(c,d) = f_\phi(x)$ and also take a second degraded image $y'$ from a different clean image under the same degradation. If the information-separation claim is correct, $f_\phi^{-1}(y', d)$ should look like the second image in coarse structure while retaining fine details from the first; if it instead shows blur, color bleed, or texture from the wrong image, the coarse/detail split has not isolated the degradation. Quantify this with PSNR and LPIPS against the ideal second image and against a no-guidance diffusion baseline.

Watch

Extended reading notes

Core claim

The central claim is that the perfect-reconstruction property of an invertible neural network can be turned into a degradation simulator and, at the same time, into an implicit inverse operator for guiding diffusion sampling. Writing the forward transform as $[c,d] = f_\phi(x)$ and training it with $\mathcal{L}(\phi) = \frac{1}{N}\sum_i \|f^c_\phi(x_i) - y_i\|^2$ makes the coarse output $c$ approximate the degraded measurement; the authors argue that, because the transform is invertible, the detail output $d$ then necessarily carries the information lost in degradation. At each reverse step they form $\hat{x}_{0,t} = f_\phi^{-1}(y, d_t)$ and update the sample by a gradient step on $\|\hat{x}_{0,t} - x_{0,t}\|^2$. The blind variant conditions the INN on an implicit degradation embedding extracted by a pretrained estimator, refines the INN parameters during sampling against the observed measurement, and starts from a diffusion-initialized clean estimate to cut the number of timesteps.

Load-bearing premise

The load-bearing premise is that forcing the INN's coarse channel to equal the degraded image automatically forces its detail channel to hold exactly the information lost in degradation, so that substituting the measurement and inverting yields a trustworthy intermediate restoration rather than a distorted or off-manifold image.

Editorial extensions

If this is right

  • Non-blind restoration no longer requires an analytic expression of the degradation operator: any degradation that can be represented by paired training data can be simulated by the INN.
  • Blind restoration can adapt to an unseen degradation at test time by finetuning the INN against the observed measurement during sampling.
  • The pretrained diffusion model is left unchanged, so its generative prior is preserved and only the lightweight INN (about 0.71M parameters for INDIGO, 0.91M for BlindINDIGO) is trained.
  • Combining a diffusion-initialized starting point with DDIM sampling reduces the required number of function evaluations from 1000 to 250 with only a modest perceptual-quality loss.
  • Because the guidance is a gradient step on the distance between the INN-refined image and the diffusion estimate, the procedure inherits the posterior-sampling interpretation of earlier gradient-based diffusion solvers.

Reading between the lines

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

  • An implication the authors leave implicit is that any inverse problem with paired data, including medical or scientific imaging where the forward model is non-linear and not analytically available, could use the same INN-as-degradation-simulator mechanism.
  • The information-separation assumption could be probed directly: on a known synthetic degradation, compare the detail channel $d$ with the actual lost high-frequency content and check whether inverting with a different measurement preserves the expected structure.
  • The consensus strategy, which averages guidance over parallel diffusion chains, suggests that variance reduction across chains is a cheap robustness lever at high noise; a theory of why three chains outperform two or four would be a natural follow-up.
  • A testable extension is to replace the fixed gradient scale $\zeta$ with a noise-level-dependent schedule, since the paper's own ablation shows that too-small $\zeta$ weakens data consistency while too-large $\zeta$ amplifies noise.
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 / 5 minor

Summary. The paper proposes two inference-time algorithms for image restoration with pretrained diffusion models. INDIGO trains a lifting-inspired invertible neural network so that the coarse output of its forward transform matches the degraded measurement; at sampling time it replaces the coarse channel of the intermediate estimate by the measurement and inverts the INN to produce a data-consistent estimate that guides the next reverse-diffusion step. BlindINDIGO extends this to unknown degradations with a conditional INN, a degradation-embedding extractor, an initialization from a pretrained restoration network, and test-time finetuning of the INN. Experiments report competitive or state-of-the-art numbers against ILVR, DDRM, DPS, PGDiff, DifFace, DR2, and StableSR on FFHQ, CelebA-HQ, ImageNet, DRealSR, and CelebChild, with ablations on step size, loss design, initialization, finetuning, sampling acceleration, and a consensus strategy.

Significance. If the central mechanism works, the contribution is practically significant: it removes the need for an analytic degradation model in non-blind restoration, treats nonlinear and real-world degradations through learned simulation, and leaves the pretrained diffusion model untouched. The INN is light (0.71M and 0.91M parameters), the inversion is exact by construction, and the experiments are broad, with component ablations and a DDIM variant that reduces runtime. The main conceptual risk is that the method's success depends on an information-separation property that is asserted rather than established. Since the paper is empirical, this can be addressed with targeted experiments, but the property is load-bearing for both INDIGO and BlindINDIGO.

major comments (3)
  1. [Section III-A, Eq. (12), Algorithm 1] The load-bearing claim that minimizing ||f^c_phi(x_i) - y_i||^2 forces the detail channel f^d_phi(x_i) to contain exactly the lost information is not justified by invertibility alone. Invertibility only makes (c,d)=f_phi(x) a bijective reparameterization; the loss contains no term on d, so training can allocate information between c and d arbitrarily as long as c matches y on the training pairs. Because the guidance step in Algorithm 1 replaces c_t by y and inverts with the diffusion-estimated d_t, an incorrect allocation could distort x0,t or impose consistency with a surrogate degradation different from the true H. The evidence in Fig. 18, showing that the coarse output tracks degradation level, does not test the required separation. Please add a direct disentanglement evaluation: measure the residual fidelity of f^{-1}(y, f^d(x)) on validation data, check whether d_t changes with y for fixed x, and report a final data-consistency metric such as ||H(f^{-1}(y,d_t)) - y||_2 for the restored images.
  2. [Algorithm 2, Eq. (16), Table VII, Fig. 17] The test-time finetuning update at the end of each iteration, phi <- phi - l * grad_phi ||c_t - y||^2, uses c_t = f^c_phi(x0,t, gamma_deg), where x0,t is itself produced by the previous guidance using the same phi. This self-referential loop has no convergence or stability guarantee and can reinforce estimation errors rather than correct the degradation model. The manuscript's own Fig. 17 shows that the pretrained INN can fail badly on complex real-world degradation, and the ablation in Table VII (cases 6 vs 7 and 9 vs 10) reports only aggregate improvement. Please add an analysis of the finetuning trajectory (e.g., loss and output quality over iterations), a check that phi does not drift arbitrarily far from the pretrained initialization, and a failure-mode discussion for the cases where finetuning does not help.
  3. [Section IV-A3 and IV-A4, Figs. 8 and 9] The claims that INDIGO handles nonlinear JPEG-degradation and real-world DRealSR degradation are supported only by visual examples and no quantitative comparison or consistency metric. Since the non-blind contribution is specifically the ability to work without an analytic degradation model, the absence of any baseline (e.g., a diffusion method using a differentiable JPEG approximation, or a quantitative comparison against existing methods on DRealSR) leaves the practical advantage unquantified. Please add quantitative results for these settings and, where possible, compare with methods that can handle the same degradation or report a data-consistency measure.
minor comments (5)
  1. [Table IV] StableSR is cited as [1] in the table but as [4] in the text; the reference numbering should be corrected.
  2. [Eq. (16) and Algorithm 2] The loss weights lambda_F and lambda_I and the definitions of LF and LI are introduced after Algorithm 2; define them before the algorithm is presented.
  3. [Fig. 18] The caption says 'different conditions' but the experiment varies degradation level; the wording should match the actual setup.
  4. [Appendix A and Table I] The consensus strategy triples the number of sampling trajectories, but the runtime cost is not reported for Table I; state clearly that 'Ours' in Table I uses the consensus strategy and report the corresponding runtime so the comparison is transparent.
  5. [General] No code is released; for reproducibility, release code and trained models or provide full training details for the INN (optimizer, epochs, batch size, data augmentation).

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the INN is trained on independent pairs and the data-consistency step is an imposed constraint; the central claim is empirically evaluated, not reduced to its inputs.

full rationale

The paper's central workflow is algorithmic and empirically evaluated, not a derivation whose outputs equal its inputs by construction. The INN in Eq. (12) is trained on independent pairs (x_i, y_i) to match the coarse channel to the degraded image; the test-time operation \hat{x}_{0,t}=f_phi^{-1}(y,d_t) in Algorithm 1 is an imposed data-consistency projection, not a fitted constant renamed as a prediction. The final result is produced by a pretrained diffusion sampler plus a gradient step, and the quantitative comparisons are on held-out validation/test sets, so successes and failures are externally falsifiable. The assertion in Sec. III-A that 'due to invertibility, the other part will inevitably represent the detailed information lost' is an unproven disentanglement assumption and a correctness risk, but it is not a circular reduction: no equation forces the final image to equal the training target or the measurement. The self-citations to the authors' MMSP paper [24] and the co-authored LINN architecture [37] are not load-bearing; the algorithm is described self-containedly in Algorithms 1-3. The blind test-time finetuning (phi update in Algorithm 2) adapts the INN to y and could in principle reinforce estimation errors, but it is a joint estimation procedure rather than a case of fitting a parameter to a subset and then predicting that same subset. Overall, no step reduces to its own input by definition.

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

The central claim rests on a trained INN simulator, a pretrained DEM and diffusion model, and several hand-tuned hyperparameters. No new physical or conceptual entities are introduced beyond the learned degradation embedding and the coarse/detail decomposition, and no independent falsifiable handles are proposed.

free parameters (6)
  • Gradient scale zeta = 0.5 (INDIGO), 1.5 (BlindINDIGO)
    Chosen by hand in Section IV-C4 based on the visual tradeoff between measurement consistency and noise amplification.
  • Loss weights lambda_F, lambda_I = 2.5, 1.0
    Set empirically in Section III-C to balance measurement-space and image-space losses in the blind case.
  • Starting timestep N (blind) = 400
    Set in Section IV-B1 to start from a forward-diffused initial estimate, trading off speed and reconstruction quality.
  • Finetuning learning rate l = 1e-5
    Set in Section IV-B1 for test-time adaptation of the INN parameters.
  • Number of consensus paths = 3 (non-blind), 1 (blind)
    Chosen by ablation in Appendix A, where three parallel paths give the best PSNR gain in the non-blind case but double the runtime in the blind case.
  • INN architecture choices = 2 levels, 4 PNet/UNet pairs, 32 feature channels, filter size 5, gamma dimension 128
    Hand-designed following the lifting-inspired architecture of [37], with details in Sections IV-A1 and IV-B1; these choices affect capacity and are not derived.
assumptions (5)
  • standard math The lifting-inspired INN is perfectly invertible by construction.
    Invertibility follows from the lifting scheme in Section II-B (Eqs. 5-10), and the paper relies on it for the exact reconstruction step.
  • domain assumption Pre-trained unconditional diffusion models provide a valid image prior for FFHQ and ImageNet.
    The method assumes off-the-shelf DDPM/DDIM checkpoints are expressive enough for the restoration tasks; these are cited from [8], [10], and [22].
  • domain assumption The pre-trained degradation estimation module (DEM) produces a degradation embedding that generalizes to unseen degradations.
    BlindINDIGO conditions the INN on the embedding from DEM [53], but no evidence is provided that the embedding covers the real-world degradations tested.
  • ad hoc to paper Minimizing L(phi) in Eq. (12) makes the coarse output match the degraded image and forces the detail output to carry the lost information.
    This separation hypothesis is asserted in Sections III-A and III-B as an inevitable consequence of invertibility, but it is not proven and the entire data-consistency step depends on it.
  • domain assumption A fixed-step gradient update on the consistency loss keeps the sample on the natural-image manifold.
    The gradient guidance in Algorithms 1 and 2 is a heuristic; no theoretical guarantee is given that the updated x_{t-1} remains a valid diffusion sample.

how reviews work

0 comments
Cite this review

Pith. "Pith review of INDIGO+: A Unified INN-Guided Probabilistic Diffusion Algorithm for Blind and Non-Blind Image Restoration." pith.science (2026). https://pith.science/paper/IYDMQVLY

@misc{pith2026250114014,
  author       = {Pith},
  title        = {Pith review of: INDIGO+: A Unified INN-Guided Probabilistic Diffusion Algorithm for Blind and Non-Blind Image Restoration},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/IYDMQVLY}},
  note         = {Machine review of arXiv:2501.14014}
}
read the original abstract

Generative diffusion models are becoming one of the most popular prior in image restoration (IR) tasks due to their remarkable ability to generate realistic natural images. Despite achieving satisfactory results, IR methods based on diffusion models present several limitations. First of all, most non-blind approaches require an analytical expression of the degradation model to guide the sampling process. Secondly, most existing blind approaches rely on families of pre-defined degradation models for training their deep networks. The above issues limit the flexibility of these approaches and so their ability to handle real-world degradation tasks. In this paper, we propose a novel INN-guided probabilistic diffusion algorithm for non-blind and blind image restoration, namely INDIGO and BlindINDIGO, which combines the merits of the perfect reconstruction property of invertible neural networks (INN) with the strong generative capabilities of pre-trained diffusion models. Specifically, we train the forward process of the INN to simulate an arbitrary degradation process and use the inverse to obtain an intermediate image that we use to guide the reverse diffusion sampling process through a gradient step. We also introduce an initialization strategy, to further improve the performance and inference speed of our algorithm. Experiments demonstrate that our algorithm obtains competitive results compared with recently leading methods both quantitatively and visually on synthetic and real-world low-quality images.

Figures

Figures reproduced from arXiv: 2501.14014 by the authors.

Figure 1
Figure 1. Comparisons with state-of-the-art blind image restoration approaches [1]–[4] on the real-world low-quality images. Our [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. The wavelet transform obtained using the lifting [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Overview of our INDIGO for non-blind image restora [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (16 more)
Figure 4
Figure 4. Figure 4: Overview of our BlindINDIGO for blind image restoration. Given a degraded image [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: The forward and inverse transform of our INN during [PITH_FULL_IMAGE:figures/full_fig_p005_5.png]
Figure 6
Figure 6. Figure 6: (a) The architecture of the PNet/UNet. (b)The archi [PITH_FULL_IMAGE:figures/full_fig_p005_6.png]
Figure 7
Figure 7. Figure 7: Comparisons with state-of-the-art image restoration ap [PITH_FULL_IMAGE:figures/full_fig_p006_7.png]
Figure 8
Figure 8. Figure 8: Results of our algorithm on solving the non-blind inverse problem with Jpeg compression on CelabA HQ vali￾dation dataset. of the following loss: Ltotal = λF LF (ct, y) + λILI (xˆ0,t, x0,t), (16) where we constrain the measurement space with LF (ct, y) and the high-qual…
Figure 9
Figure 9. Figure 9: Result of our algorithm on reconstructing real images from DRealSR [44] with resolution enhancement by a factor 4 per direction (non-blind). x0,t, the parameters of our INN, fϕ, are updated to reduce LF (ct, y) in Eq. (16). Accelerating our algorithm with initializatio…
Figure 10
Figure 10. Figure 10: Comparisons on 4x blind SR with mild degradation on CelebA-HQ. (a) Input (b) DifFace (c) PGDiff (d) DR2 (e) StableSR (f) Ours (g) GT [PITH_FULL_IMAGE:figures/full_fig_p008_10.png]
Figure 11
Figure 11. Figure 11: Comparisons on 4x blind SR with medium degradation on CelebA-HQ. (a) Input (b) DifFace (c) PGDiff (d) DR2 (e) StableSR (f) Ours (g) GT [PITH_FULL_IMAGE:figures/full_fig_p008_11.png]
Figure 12
Figure 12. Figure 12: Comparisons on 4x blind SR with severe degradation on CelebA-HQ [PITH_FULL_IMAGE:figures/full_fig_p008_12.png]
Figure 13
Figure 13. Figure 13: Result of our BlindINDIGO on 4x super-resolution on the ImageNet dataset. [PITH_FULL_IMAGE:figures/full_fig_p010_13.png]
Figure 14
Figure 14. Figure 14: Comparisons with blind state-of-the-art image restoration approaches [1]–[4] on real-world dataset Celaba-Child. [PITH_FULL_IMAGE:figures/full_fig_p010_14.png]
Figure 15
Figure 15. Figure 15: Sampling process of our approach. in the noiseless case and outperforms significantly the other methods in noisy settings [PITH_FULL_IMAGE:figures/full_fig_p010_15.png]
Figure 16
Figure 16. Figure 16: Ablation study on the choice of step size schedule for our INDIGO. [PITH_FULL_IMAGE:figures/full_fig_p012_16.png]
Figure 17
Figure 17. Figure 17: Some examples where our algorithm with the [PITH_FULL_IMAGE:figures/full_fig_p012_17.png]
Figure 18
Figure 18. Figure 18: Effect of our INN. Low-Resolution T=250 T=1000 Ground Truth [PITH_FULL_IMAGE:figures/full_fig_p013_18.png]
Figure 19
Figure 19. Figure 19: Results of our approach with T=1000 and T=250 (with [PITH_FULL_IMAGE:figures/full_fig_p013_19.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

63 extracted references · 35 canonical work pages

  1. [1]

    Dr2: Diffusion-based robust degradation remover for blind face restoration,

    Z. Wang, Z. Zhang, X. Zhang, H. Zheng, M. Zhou, Y . Zhang, and Y . Wang, “Dr2: Diffusion-based robust degradation remover for blind face restoration,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2023, pp. 1704–1713

    work page 2023

  2. [2]

    Difface: Blind face restoration with diffused error contraction,

    Z. Yue and C. C. Loy, “Difface: Blind face restoration with diffused error contraction,” arXiv preprint arXiv:2212.06512 , 2022

    arXiv 2022

  3. [3]

    PGDiff: Guiding diffusion models for versatile face restoration via partial guidance,

    P. Yang, S. Zhou, Q. Tao, and C. C. Loy, “PGDiff: Guiding diffusion models for versatile face restoration via partial guidance,” in NeurIPS, 2023

    work page 2023

  4. [4]

    Exploiting diffusion prior for real-world image super-resolution,

    J. Wang, Z. Yue, S. Zhou, K. C. Chan, and C. C. Loy, “Exploiting diffusion prior for real-world image super-resolution,” arXiv preprint arXiv:2305.07015, 2023

    arXiv 2023

  5. [5]

    Deep learning techniques for inverse problems in imaging,

    G. Ongie, A. Jalal, C. A. Metzler, R. G. Baraniuk, A. G. Dimakis, and R. Willett, “Deep learning techniques for inverse problems in imaging,” IEEE Journal on Selected Areas in Information Theory , vol. 1, no. 1, pp. 39–56, 2020

    2020

  6. [6]

    Plug-and- play priors for model based reconstruction,

    S. V . Venkatakrishnan, C. A. Bouman, and B. Wohlberg, “Plug-and- play priors for model based reconstruction,” in Proc. of IEEE Global Conference on Signal and Information Processing (GlobalSIP) , 2013, pp. 945–948

    work page 2013

  7. [7]

    Plug-and-play methods for integrating physical and learned models in computational imaging: Theory, algorithms, and applications,

    U. S. Kamilov, C. A. Bouman, G. T. Buzzard, and B. Wohlberg, “Plug-and-play methods for integrating physical and learned models in computational imaging: Theory, algorithms, and applications,” IEEE Signal Processing Magazine , vol. 40, no. 1, pp. 85–97, 2023

    work page 2023

  8. [8]

    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

Show all 63 references
  1. [9]

    Denoising diffusion implicit models,

    J. Song, C. Meng, and S. Ermon, “Denoising diffusion implicit models,” in International Conference on Learning Representations (ICLR) , 2021

    2021

  2. [10]

    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. [11]

    Generative modeling by estimating gradients of the data distribution,

    Y . Song and S. Ermon, “Generative modeling by estimating gradients of the data distribution,” in Advances in Neural Information Processing Systems, 2019, pp. 11 895–11 907

    2019

  4. [12]

    Score-based generative modeling through stochastic differen- tial equations,

    Y . Song, J. Sohl-Dickstein, D. P. Kingma, A. Kumar, S. Ermon, and B. Poole, “Score-based generative modeling through stochastic differen- tial equations,” in International Conference on Learning Representations (ICLR), 2022

    2022

  5. [13]

    Ilvr: Conditioning method for denoising diffusion probabilistic models,

    J. Choi, S. Kim, Y . Jeong, Y . Gwon, and S. Yoon, “Ilvr: Conditioning method for denoising diffusion probabilistic models,” in International Conference on Computer Vision (ICCV) . IEEE, 2021, pp. 14 347– 14 356

    2021

  6. [14]

    Come-closer-diffuse-faster: Accelerat- ing conditional diffusion models for inverse problems through stochastic contraction,

    H. Chung, B. Sim, and J. C. Ye, “Come-closer-diffuse-faster: Accelerat- ing conditional diffusion models for inverse problems through stochastic contraction,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2022, pp. 12 413–12 422

    2022

  7. [15]

    Repaint: Inpainting using denoising diffusion probabilistic models,

    A. Lugmayr, M. Danelljan, A. Romero, F. Yu, R. Timofte, and L. Van Gool, “Repaint: Inpainting using denoising diffusion probabilistic models,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , 2022, pp. 11 461–11 471

    2022

  8. [16]

    Solving linear inverse problems using the prior implicit in a denoiser,

    Z. Kadkhodaie and E. P. Simoncelli, “Solving linear inverse problems using the prior implicit in a denoiser,” in NeurIPS 2020 Workshop on Deep Learning and Inverse Problems

    2020

  9. [17]

    SNIPS: Solving noisy inverse problems stochastically,

    B. Kawar, G. Vaksman, and M. Elad, “SNIPS: Solving noisy inverse problems stochastically,” Advances in Neural Information Processing Systems, vol. 34, pp. 21 757–21 769, 2021

    2021

  10. [18]

    Denoising diffusion restoration models,

    B. Kawar, M. Elad, S. Ermon, and J. Song, “Denoising diffusion restoration models,” in Advances in Neural Information Processing Systems, 2022

    2022

  11. [19]

    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,” in International Conference on Learning Representations (ICLR) , 2023

    2023

  12. [20]

    Solving inverse problems in medical imaging with score-based generative models,

    Y . Song, L. Shen, L. Xing, and S. Ermon, “Solving inverse problems in medical imaging with score-based generative models,” arXiv preprint arXiv:2111.08005, 2021

    2021 arXiv

  13. [21]

    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 (ICLR) , 2023

    2023

  14. [22]

    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,” in International Conference on Learning Representations (ICLR) , 2023

    2023

  15. [23]

    DOLCE: A model-based probabilistic diffusion framework for limited-angle ct reconstruction,

    J. Liu, R. Anirudh, J. J. Thiagarajan, S. He, K. A. Mohan, U. S. Kamilov, and H. Kim, “DOLCE: A model-based probabilistic diffusion framework for limited-angle ct reconstruction,” arXiv preprint arXiv:2211.12340 , 2022

    2022 arXiv

  16. [24]

    Indigo: An inn-guided proba- bilistic diffusion algorithm for inverse problems,

    D. You, A. Floros, and P. L. Dragotti, “Indigo: An inn-guided proba- bilistic diffusion algorithm for inverse problems,” in 2023 IEEE 25th International Workshop on Multimedia Signal Processing (MMSP) , 2023, pp. 1–6

    2023

  17. [25]

    Parallel diffusion models of operator and image for blind inverse problems,

    H. Chung, J. Kim, S. Kim, and J. C. Ye, “Parallel diffusion models of operator and image for blind inverse problems,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2023, pp. 6059–6069

    2023

  18. [26]

    Gibbsddrm: A partially collapsed gibbs sampler for solving blind inverse problems with denoising diffusion restoration,

    N. Murata, K. Saito, C.-H. Lai, Y . Takida, T. Uesaka, Y . Mitsufuji, and S. Ermon, “Gibbsddrm: A partially collapsed gibbs sampler for solving blind inverse problems with denoising diffusion restoration,” arXiv preprint arXiv:2301.12686 , 2023

    2023 arXiv

  19. [27]

    Generative diffusion prior for unified image restoration and enhancement,

    B. Fei, Z. Lyu, L. Pan, J. Zhang, W. Yang, T. Luo, B. Zhang, and B. Dai, “Generative diffusion prior for unified image restoration and enhancement,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2023, pp. 9935–9946

    2023

  20. [28]

    High- resolution image synthesis with latent diffusion models,

    R. Rombach, A. Blattmann, D. Lorenz, P. Esser, and B. Ommer, “High- resolution image synthesis with latent diffusion models,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recogni- tion (CVPR), 2022, pp. 10 684–10 695

    2022

  21. [29]

    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

  22. [30]

    Classifier-free diffusion guidance,

    J. Ho and T. Salimans, “Classifier-free diffusion guidance,” arXiv preprint arXiv:2207.12598, 2022

    2022 arXiv

  23. [31]

    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 , 2022

    2022

  24. [32]

    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

  25. [33]

    Denoising diffusion probabilistic models for robust image super-resolution in the wild,

    H. Sahak, D. Watson, C. Saharia, and D. Fleet, “Denoising diffusion probabilistic models for robust image super-resolution in the wild,”arXiv preprint arXiv:2302.07864, 2023

    2023 arXiv

  26. [34]

    Srdiff: Single image super-resolution with diffusion probabilistic mod- els,

    H. Li, Y . Yang, M. Chang, S. Chen, H. Feng, Z. Xu, Q. Li, and Y . Chen, “Srdiff: Single image super-resolution with diffusion probabilistic mod- els,” Neurocomputing, vol. 479, pp. 47–59, 2022

    2022

  27. [35]

    Diffbfr: Bootstrapping diffusion model towards blind face restoration,

    X. Qiu, C. Han, Z. Zhang, B. Li, T. Guo, and X. Nie, “Diffbfr: Bootstrapping diffusion model towards blind face restoration,” arXiv preprint arXiv:2305.04517, 2023

    2023 arXiv

  28. [36]

    Factoring wavelet transforms into lifting steps,

    I. Daubechies and W. Sweldens, “Factoring wavelet transforms into lifting steps,” Journal of Fourier analysis and applications, vol. 4, no. 3, pp. 247–269, 1998

    1998

  29. [37]

    WINNet: Wavelet-inspired invertible network for image denoising,

    J.-J. Huang and P. L. Dragotti, “WINNet: Wavelet-inspired invertible network for image denoising,” IEEE Transactions on Image Processing, vol. 31, pp. 4377–4392, 2022

    2022

  30. [38]

    Nice: Non-linear independent components estimation,

    L. Dinh, D. Krueger, and Y . Bengio, “Nice: Non-linear independent components estimation,” arXiv preprint arXiv:1410.8516 , 2014

    2014 arXiv

  31. [39]

    Density estimation using Real NVP,

    L. Dinh, J. Sohl-Dickstein, and S. Bengio, “Density estimation using Real NVP,” arXiv preprint arXiv:1605.08803 , 2016

    2016 arXiv

  32. [40]

    The reversible residual network: Backpropagation without storing activations,

    A. N. Gomez, M. Ren, R. Urtasun, and R. B. Grosse, “The reversible residual network: Backpropagation without storing activations,” arXiv preprint arXiv:1707.04585, 2017

    2017 arXiv

  33. [41]

    i-RevNet: Deep invertible networks,

    J.-H. Jacobsen, A. W. Smeulders, and E. Oyallon, “i-RevNet: Deep invertible networks,” in International Conference on Learning Repre- sentations, 2018

    2018

  34. [42]

    Interactive multi-dimension modulation with dynamic controllable residual learning for image restoration,

    J. He, C. Dong, and Y . Qiao, “Interactive multi-dimension modulation with dynamic controllable residual learning for image restoration,” in Computer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part XX 16 . Springer, 2020, pp. 53–68

    2020

  35. [43]

    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

  36. [44]

    Component divide-and-conquer for real-world image super-resolution,

    P. Wei, Z. Xie, H. Lu, Z. Zhan, Q. Ye, W. Zuo, and L. Lin, “Component divide-and-conquer for real-world image super-resolution,” in Proceed- ings of the European Conference on Computer Vision (ECCV) , 2020

    2020

  37. [45]

    Very deep convolutional networks for large-scale image recognition,

    K. Simonyan and A. Zisserman, “Very deep convolutional networks for large-scale image recognition,” arXiv preprint arXiv:1409.1556 , 2014

    2014 arXiv

  38. [46]

    Swinir: Image restoration using swin transformer,

    J. Liang, J. Cao, G. Sun, K. Zhang, L. Van Gool, and R. Timofte, “Swinir: Image restoration using swin transformer,” in Proceedings of the IEEE/CVF international conference on computer vision , 2021, pp. 1833–1844. JOURNAL OF LATEX CLASS FILES, VOL. 14, NO. 8, AUGUST 2021 15

    2021

  39. [47]

    A style-based generator architecture for generative adversarial networks,

    T. Karras, S. Laine, and T. Aila, “A style-based generator architecture for generative adversarial networks,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2019, pp. 4401–4410

    2019

  40. [48]

    Progressive growing of gans for improved quality, stability, and variation,

    T. Karras, T. Aila, S. Laine, and J. Lehtinen, “Progressive growing of gans for improved quality, stability, and variation,” in International Conference on Learning Representations (ICLR) , 2018

    2018

  41. [49]

    Ima- geNet: A large-scale hierarchical image database,

    J. Deng, W. Dong, R. Socher, L.-J. Li, K. Li, and L. Fei-Fei, “Ima- geNet: A large-scale hierarchical image database,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2009, pp. 248–255

    2009

  42. [50]

    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

  43. [51]

    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 Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2018, pp. 586–595

    2018

  44. [52]

    Making a “completely blind

    A. Mittal, R. Soundararajan, and A. C. Bovik, “Making a “completely blind” image quality analyzer,” IEEE Signal processing letters , vol. 20, no. 3, pp. 209–212, 2012

    2012

  45. [53]

    Knowledge distillation based degradation estimation for blind super-resolution,

    B. Xia, Y . Zhang, Y . Wang, Y . Tian, W. Yang, R. Timofte, and L. Van Gool, “Knowledge distillation based degradation estimation for blind super-resolution,” arXiv preprint arXiv:2211.16928 , 2022

    2022 arXiv

  46. [54]

    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

  47. [55]

    Ntire 2017 challenge on single image super-resolution: Methods and results,

    R. Timofte, E. Agustsson, L. Van Gool, M.-H. Yang, and L. Zhang, “Ntire 2017 challenge on single image super-resolution: Methods and results,” in Proceedings of the IEEE conference on computer vision and pattern recognition workshops , 2017, pp. 114–125

    2017

  48. [56]

    Learning spatial attention for face super-resolution,

    C. Chen, D. Gong, H. Wang, Z. Li, and K.-Y . K. Wong, “Learning spatial attention for face super-resolution,” IEEE Transactions on Image Processing, vol. 30, pp. 1219–1231, 2020

    2020

  49. [57]

    Gan prior embedded network for blind face restoration in the wild,

    T. Yang, P. Ren, X. Xie, and L. Zhang, “Gan prior embedded network for blind face restoration in the wild,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2021, pp. 672– 681

    2021

  50. [58]

    Towards real-world blind face restoration with generative facial prior,

    X. Wang, Y . Li, H. Zhang, and Y . Shan, “Towards real-world blind face restoration with generative facial prior,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2021, pp. 9168–9178

    2021

  51. [59]

    Towards robust blind face restoration with codebook lookup transformer,

    S. Zhou, K. C. Chan, C. Li, and C. C. Loy, “Towards robust blind face restoration with codebook lookup transformer,” in NeurIPS, 2022

    2022

  52. [60]

    Vqfr: Blind face restoration with vector-quantized dictionary and parallel decoder,

    Y . Gu, X. Wang, L. Xie, C. Dong, G. Li, Y . Shan, and M.-M. Cheng, “Vqfr: Blind face restoration with vector-quantized dictionary and parallel decoder,” in ECCV, 2022

    2022

  53. [61]

    No-reference image quality assessment in the spatial domain,

    A. Mittal, A. K. Moorthy, and A. C. Bovik, “No-reference image quality assessment in the spatial domain,” IEEE Transactions on image processing, vol. 21, no. 12, pp. 4695–4708, 2012

    2012

  54. [62]

    Arcface: Additive angular margin loss for deep face recognition,

    J. Deng, J. Guo, N. Xue, and S. Zafeiriou, “Arcface: Additive angular margin loss for deep face recognition,” in Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2019, pp. 4690– 4699

    2019

  55. [63]

    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. Di You (Student Member, IEEE) received the bach- elor’s degree in electronic and information engineer- ing from D...

    2023 arXiv

Pith tools

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