Pith. sign in

REVIEW 4 major objections 6 minor 55 references

DR-BFR: Degradation Representation with Diffusion Models for Blind Face Restoration

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

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

Pith's one-line read DR-BFR claims that conditioning a latent diffusion restoration model on a content-independent degradation representation, learned from contrastive learning and low-quality reconstruction, yields more natural and faithful face restoration…

desk verdict A genuinely new degradation-prompt conditioning mechanism for diffusion-based blind face restoration, with strong FID evidence on most datasets but a direct NIQE overclaim in the text and a load-bearing patch-homogeneity assumption that real-world degradations will violate. read the letter →

arxiv 2411.10508 v1 pith:KHQCGHBD submitted 2024-11-15 cs.CV

Xinmin Qiu , Bonan Li , Zicheng Zhang , Congying Han , Tiande Guo This is my paper · ORCID
classification cs.CV
keywords blindfacerestorationdegradationrepresentationcontrastivelearninglatentdiffusionmodelpromptimage
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper sets out to show that blind face restoration fails when the diffusion model must guess the degradation, and that this can be fixed by decoupling a degradation prompt from the low-quality face. It introduces DR-BFR, in which a Degradation Representation Module learns a content-independent vector summarising the blur, noise, downsampling, and compression in an LQ face, using patch-level contrastive learning plus reconstruction of the LQ image. A Latent Diffusion Restoration Module then receives this vector alongside LQ content features and uses cross-attention with a step-dependent coefficient to denoise. The authors report the best FID among all compared methods on the synthetic CelebA-Test and on all four real-world face datasets, arguing that the decoupled prompt makes restoration more natural and more faithful than prior GAN-, codebook-, or diffusion-based approaches.

What carries the argument

The machinery is the Degradation Representation Module (DRM) coupled to a Latent Diffusion Restoration Module. DRM uses a ConvNet encoder similar to VGG-19, contrastive loss on patches, an AdaIN U-Net generator $G_{LQ}$ that reconstructs LQ from DR and HQ, and a Gaussian distribution loss. The reconstruction loss makes DR carry all degradation information; the contrastive loss makes it content-independent; the distribution loss regularizes the DR space. The LDRM is a latent diffusion model whose U-Net takes concatenated $f_{LQ}$ and uses cross-attention on $f_{DR}$ with per-step MLP coefficients, so the degradation prompt is weighted differently at different denoising stages.

What would settle it

Construct a synthetic LQ face whose left half is heavily blurred and right half is noisy, or take a real photo with a watermark; if DR-BFR cannot restore it without residual artifacts (as the paper notes happens with watermarks), or if the DR from the two halves of the same image are pulled together despite representing different degradations, the central decoupling claim is falsified.

Watch

Extended reading notes

Core claim

The central claim is that degrading a face and decoupling the degradation from its content yields a compact, content-free representation that works as a downstream prompt for diffusion-based restoration. DRM is trained so that two random patches of the same LQ image produce positive samples of each other, while patches from different images are negatives; a U-Net generator with AdaIN takes the DR plus the HQ image to reconstruct the LQ image, and a distribution loss pushes the DR manifold toward a Gaussian. This DR, denoted $f_{DR}$, is inserted into the LDM's cross-attention along with the LQ content feature $f_{LQ}$, with a scaling coefficient $\alpha_t$ produced by an MLP from the timestep $t$. On CelebA-Test, DR-BFR achieves FID 13.686, NIQE 5.0113, and LPIPS 0.2499, the best among the compared methods, and it leads in FID on LFW, CelebChild, WebPhoto-Test, and Wider.

Load-bearing premise

The load-bearing premise is that different patches of the same low-quality face share the same degradation, so they can serve as positive pairs in contrastive learning; when real degradations are spatially uneven, such as watermarks or local blur, that premise fails and the learned representation can conflate content with artifacts.

Editorial extensions

If this is right

  • Conditioning a diffusion restoration model on a decoupled degradation prompt delivers the best perceptual quality among the compared methods: FID 13.686 on CelebA-Test, NIQE 5.0113, and the best FID on LFW, CelebChild, WebPhoto-Test, and Wider.
  • Making the degradation prompt's weight depend on the denoising step via cross-attention lets the model apply degradation cues at the right stages, improving both naturalness and fidelity.
  • The decoupled DR transfers from synthetic training degradations to real-world photos, because the method's real-world results improve without paired HQ data at inference.
  • Each ingredient in DRM matters: removing contrastive learning or LQ reconstruction raises FID to 18.730 or 19.644, showing both losses are needed for a useful DR.

Reading between the lines

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

  • In the editor's reading, the same content-independent degradation-vector recipe should transfer to other blind inverse problems, since the DR is task-agnostic and the LDM backbone can be swapped.
  • Because the DR is content-independent and regularized toward a Gaussian, one could edit it directly to control restoration strength, e.g., removing a watermark or tuning blur without retraining.
  • The paper's residual-watermark failure suggests a concrete test: augment DRM training with unpaired real degraded images, or replace image-level contrastive positives with patch-level clustering by local degradation type.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. The paper proposes DR-BFR, a blind face restoration method built on a latent diffusion model. A Degradation Representation Module (DRM) is trained with patch-based contrastive learning, an LQ reconstruction loss, and an energy-distance distribution loss to produce a content-independent degradation representation f_DR from an LQ face. A Latent Diffusion Restoration Module (LDRM) then conditions an LDM on both VAE content features and f_DR, injecting the degradation representation through a step-dependent cross-attention mechanism. The authors compare DR-BFR against GAN-, codebook-, and diffusion-based baselines on CelebA-Test and four real-world datasets, and report ablations showing that removing the DR guidance degrades FID from 13.686 to 20.103 on CelebA-Test. The paper's central claim is that decoupling the degradation into a prompt improves the naturalness and generalization of diffusion-based blind face restoration.

Significance. If the reported results hold, the paper makes a useful conceptual contribution: instead of forcing the diffusion model to infer degradation implicitly, it provides an explicit, learned degradation prompt. The strongest evidence for this idea is the internal ablation in Table 3, which shows a large FID gain when the full DRM is used, and the comparison against external baselines on CelebA-Test and real-world datasets. The paper also has a clear and honest limitation statement concerning watermark artifacts. However, the significance is currently tempered by reporting inconsistencies in Table 2, by the lack of independent validation of the DR's content independence, and by the questionable use of FFHQ-FID for real-world restoration quality. These issues are local and fixable, but they currently prevent the generalization claims from being fully supported.

major comments (4)
  1. [§4.3, Table 2] Section 4.3 states that DR-BFR 'significantly outperforms the other methods in both NIQE and FID' on real-world datasets, but Table 2 does not support this statement as printed. The header lists 'FID↓ NIQE↓' for each dataset; taken literally, DR-BFR's Wider values are 4.5283 (FID) and 30.08 (NIQE), the latter being worse than DiffBIR's 25.27 and CodeFormer's 29.40. If instead the first value of each pair is NIQE, then the column headers are mislabeled, and DR-BFR's Wider FID (30.08) is worse than DiffBIR's 25.27, VQFR's 28.18, and CodeFormer's 29.40, while its LFW FID (41.67) is worse than PGDiff's 40.14. In either reading, the text overstates the results; the table and the claim need to be corrected.
  2. [§3.1, Eq. (5), §5] The patch-homogeneity premise is load-bearing for the DRM training: positive pairs in the contrastive loss are constructed from different patches of the same LQ image, justified by the claim that those patches share almost the same degradation. This holds for the spatially uniform synthetic degradation of Eq. (11), but not for spatially varying real-world degradations such as watermarks, local motion blur, or uneven noise. The limitation section already concedes that watermark inputs retain residual artifacts, which is exactly the predicted failure mode of this premise. Because the paper's real-world generalization claim rests mainly on FID/NIQE numbers rather than artifact-level analysis, the authors should either train or evaluate on spatially varying degradations, or explicitly delimit the applicable degradation class and soften the generalization claim.
  3. [§4.1, Fig. 6] Figure 6 is presented as evidence that the DR 'is able to represent sufficient degradation information', but the DRM is explicitly trained with the LQ reconstruction loss in Eq. (4) and Eq. (7). The figure therefore verifies that the training objective was optimized; it does not independently demonstrate that the representation is content-independent or that it generalizes to unseen degradations. An independent test would use the DR extracted from one LQ image to corrupt an unseen HQ image and compare with the true degradation applied under held-out parameters, or evaluate the frozen DR on a downstream task. The ablation in Table 3 is the strongest evidence for the method, but it does not by itself validate content independence.
  4. [§4.3, Table 2 caption] Section 4.3 says real-world results 'can only be quantitatively evaluated using non-reference metrics such as NIQE and FID', but FID is not a no-reference metric: the Table 2 caption states that FID measures the distribution gap between FFHQ and the results. A method that aggressively transforms outputs toward the FFHQ distribution can improve this FID without restoring the specific identity or degradation-related content of the input. The authors should justify why this FFHQ-FID is a meaningful fidelity measure for restoration, and should qualify the 'outperforms in FID' statements accordingly or report no-reference metrics that better reflect fidelity, such as identity similarity or artifact detection.
minor comments (6)
  1. [§3.2] The sentence 'It is difficult to extract content features that are independent of degradation, while it is relatively easier to extract content features that are independent of degradation' contains a repeated phrase and should be corrected to state the intended comparison between degradation-independent content extraction and content-independent degradation extraction.
  2. [§3.2, Eq. (10)] The notation 'CrossAttention (αt · fDR + βt, zt)' is ambiguous; the authors should specify which tensors serve as query, key, and value, and state explicitly how f_LQ is concatenated with z_t.
  3. [§4] The paper refers to an appendix for details and complete results, but the appendix is not included in the submitted manuscript; training hyperparameters, model sizes, optimizer settings, and diffusion sampling parameters are missing, which hinders reproducibility.
  4. [Table 1] The statement that lower PSNR/SSIM are acceptable because higher values 'are often observed in blurrier images' is unclear and not a substitute for a quantitative discussion of the fidelity-perceptual trade-off; consider reporting a pixel-fidelity baseline or a clearer justification.
  5. [§4.1, Fig. 7] The t-SNE visualization and cosine-similarity example in Figure 8 are qualitative; a quantitative content-independence measure, such as the variance of DR features across content variations under fixed degradation, would strengthen the claim that the DR is content-independent.
  6. [Throughout] There are several typos, including 'Quatitative' in the Table 1 caption, 'mehods' in Section 4.3, and 'reasonbly' in the Figure 1 caption; the manuscript needs a careful proofreading pass.

Circularity Check

1 steps flagged · score 1.0 of 10

No significant circularity in the central derivation: only the internal DR-reconstruction sanity check reuses its own training objective; SOTA comparisons and ablation are independent.

  1. self definitional [Section 3.1 (Eq. 4) and Section 4.1 (Figure 6)]
    "The similarity between the reconstructed image and the original one is calculated in Figure 6. The results show that highly similar LQ reconstructed images can be obtained irrespective of degradation types and random combinations of parameters, which demonstrates the maximum coverage of the degradation information by DR."

    Eq. (4) defines Lrecon = sum_i || GLQ(r2_i, f1_DRi) - p2_i ||, so the DRM is explicitly trained to reconstruct LQ patches from the DR feature plus the HQ patch. Section 4.1 then validates DR by measuring exactly this reconstruction and concluding it demonstrates maximum coverage of degradation information. The evidence is therefore the training objective itself: high similarity is enforced by construction rather than independently discovered. This circularity is peripheral; the central claim that DR improves LDM restoration is supported by external baselines and the Table 3 ablation, neither of which reduces to the DRM training loss.

full rationale

The main derivation chain of DR-BFR is not circular. The DRM is trained with contrastive learning and LQ reconstruction, and the resulting DR is then used as a cross-attention prompt in a latent diffusion model; the claimed improvement over prior methods is established by quantitative comparisons against external SOTA baselines (Tables 1 and 2) and by the ablation in Table 3. These comparisons are not fitted inputs renamed as predictions, and no load-bearing self-citation chain is used; the DR-learning idea is credited to ReDSR [19], an external citation, and the diffusion integration is tested independently. The one quasi-circular passage is the Figure 6 validation, where the DRM's reconstruction capability is demonstrated using the same objective that trained it; this is a training-convergence check, not an independent proof that DR captures degradation. The Section 5 limitation about watermark artifacts is a generalization/correctness concern about the synthetic training distribution, not a circularity in the derivation. Overall, the central empirical claim is self-contained and externally benchmarked, so circularity is negligible.

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

The central claim rests on three hand-chosen loss or sampling parameters, a synthetic degradation model, and the patch-homogeneity assumption. No new physical entities are introduced; the DR feature is a learned representation rather than an independently evidenced entity.

free parameters (3)
  • lambda_1, lambda_2
    Loss weights in Eq. (7) for contrastive learning and distribution alignment; no values are reported, yet they determine the balance between content invariance and degradation coverage.
  • tau (contrastive temperature)
    Temperature in Eq. (5); no value is reported, and this hyperparameter strongly affects how well the DR features separate by degradation type.
  • Degradation parameter ranges (sigma, r, delta, q) = sigma in [0.1,10], r in [0.8,8], delta in [0,20], q in [60,100]
    These hand-chosen ranges in Eq. (11) define the synthetic degradation distribution used for all training pairs and are not derived from real-world measurements.
assumptions (4)
  • domain assumption Patches from the same LQ image share identical degradation and differ in content (Section 3.1).
    This justifies using one patch's DR as a positive key for another patch in contrastive learning. It fails for localized degradations such as watermarks, which the paper acknowledges.
  • domain assumption The synthetic degradation model of Eq. (11) with the stated ranges covers the target real-world degradation distribution.
    Used to generate all LQ-HQ training pairs; the paper's limitations admit real-world degradations such as watermarks are not covered by this synthetic space.
  • standard math Latent Diffusion Model, VQVAE, and the DDPM formulation from Rombach et al. are correct and applicable as priors.
    The restoration module is built directly on LDM and the paper does not re-derive or verify this generative prior.
  • ad hoc to paper Aligning DR features to a Gaussian distribution (Eq. 6, Ldistribution) improves restoration performance.
    The distribution loss is introduced to ensure a reasonable arrangement of the degradation space, but no ablation isolates it and no principled justification is provided.

how reviews work

0 comments
Cite this review

Pith. "Pith review of DR-BFR: Degradation Representation with Diffusion Models for Blind Face Restoration." pith.science (2026). https://pith.science/paper/KHQCGHBD

@misc{pith2026241110508,
  author       = {Pith},
  title        = {Pith review of: DR-BFR: Degradation Representation with Diffusion Models for Blind Face Restoration},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/KHQCGHBD}},
  note         = {Machine review of arXiv:2411.10508}
}
read the original abstract

Blind face restoration (BFR) is fundamentally challenged by the extensive range of degradation types and degrees that impact model generalization. Recent advancements in diffusion models have made considerable progress in this field. Nevertheless, a critical limitation is their lack of awareness of specific degradation, leading to potential issues such as unnatural details and inaccurate textures. In this paper, we equip diffusion models with the capability to decouple various degradation as a degradation prompt from low-quality (LQ) face images via unsupervised contrastive learning with reconstruction loss, and demonstrate that this capability significantly improves performance, particularly in terms of the naturalness of the restored images. Our novel restoration scheme, named DR-BFR, guides the denoising of Latent Diffusion Models (LDM) by incorporating Degradation Representation (DR) and content features from LQ images. DR-BFR comprises two modules: 1) Degradation Representation Module (DRM): This module extracts degradation representation with content-irrelevant features from LQ faces and estimates a reasonable distribution in the degradation space through contrastive learning and a specially designed LQ reconstruction. 2) Latent Diffusion Restoration Module (LDRM): This module perceives both degradation features and content features in the latent space, enabling the restoration of high-quality images from LQ inputs. Our experiments demonstrate that the proposed DR-BFR significantly outperforms state-of-the-art methods quantitatively and qualitatively across various datasets. The DR effectively distinguishes between various degradations in blind face inverse problems and provides a reasonably powerful prompt to LDM.

Figures

Figures reproduced from arXiv: 2411.10508 by the authors.

Figure 1
Figure 1. Comparison of guidance mechanisms between previ￾ous diffusion-based methods and our DR-BFR. The key distinction lies in the approach used to handle LQ face images. In our DR￾BFR, the DR features are decoupled in advance. These decoupled features as a resonably prompt, combined with the extracted con￾tent features, serve as guiding conditions for DM. proposed DR-BFR significantly outperforms state-of-the￾art methods … view at source ↗
Figure 2
Figure 2. Framework of the proposed DR-BFR for blind face restoration task. DR-BFR is fundamentally a multi-input conditional diffu￾sion model with specific guidance. Given the LQ face, DRM decouples the content-irrelevant degradation information, while an encoder, inherited from LDRM, extracts the content of the LQ image. The diffusion-based design of DR-BFR, which incorporates degradation awareness, demonstrates superior pe… view at source ↗
Figure 3
Figure 3. Training graph of the Degradation Representation Module. Here, yˆLQ represents the LQ image generated by the re￾construction of DR and HQ image. Red brackets indicate positive samples within the same batch, while blue brackets denote nega￾tive samples. with temperature parameter τ on the LQ patches: Lcontras = X i − log qik + P i j̸=i qik − j /τ , (5) qi = MLP(f 2 DRi ), k+ i = MLP(f 1 DRi ), k− j = MLP(f 1 DRj ), T… view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: Qualitative comparisons on the CelebA-Test for blind face restoration. Our DR-BFR demonstrates strong performance in detail enhancement, hue preservation and attitude preservation compared to these latest GAN-based, dictionary-based, and diffusion-based methods. Restor…
Figure 5
Figure 5. Figure 5: Qualitative comparisons on real-world datasets. Our DR-BFR demonstrates superior performance in both detail enhancement and hue preservation, particularly on inputs with severe degradation. Zoom in for best view. Dataset Type-based LFW CelebChild WebPhoto-Test Wider Me…
Figure 6
Figure 6. Figure 6: Visualization of DR reconstruction. The output xˆ and the original LQ image x present similar degradation, indicating that DR is able to represent sufficient degradation information and reconstruct the degraded image. The bottom row represents the av￾erage PSNR/SSIM me…
Figure 7
Figure 7. Figure 7: Visualization of representations extracted from LQ images with different types of degradation. (a) illustrates repre￾sentations under randomly sampled degradation parameters, while (b) depicts representations using extreme degradation parameters across various degradat…
Figure 9
Figure 9. Figure 9: Qualitative comparisons of ablation results. Colored boxes highlight regions where local details are restored under dif￾ferent ablation schemes. Zoom in for best view. Method SSIM↑ PSNR↑ FID↓ NIQE↓ LPIPS↓ DR-None 0.6304 24.522 20.103 5.8928 0.2900 DR-CL 0.6635 25.078 1…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

55 extracted references · 28 canonical work pages

  1. [1]

    Progressive semantic- aware style transformation for blind face restoration

    Chaofeng Chen, Xiaoming Li, Lingbo Yang, Xianhui Lin, Lei Zhang, and Kwan-Yee K Wong. Progressive semantic- aware style transformation for blind face restoration. In CVPR, pages 11896–11905, 2021. 3

    work page 2021

  2. [2]

    Image super- resolution with text prompt diffusion

    Zheng Chen, Yulun Zhang, Jinjin Gu, Xin Yuan, Linghe Kong, Guihai Chen, and Xiaokang Yang. Image super- resolution with text prompt diffusion. arXiv preprint arXiv:2311.14282, 2023. 2

    arXiv 2023

  3. [3]

    Preconditioners for the stochastic training of implicit neural representations

    Shin-Fang Chng, Hemanth Saratchandran, and Simon Lucey. Preconditioners for the stochastic training of implicit neural representations. arXiv preprint arXiv:2402.08784, 2024. 3

    arXiv 2024

  4. [4]

    Inversion by di- rect iteration: An alternative to denoising diffusion for image restoration

    Mauricio Delbracio and Peyman Milanfar. Inversion by di- rect iteration: An alternative to denoising diffusion for image restoration. arXiv preprint arXiv:2303.11435, 2023. 3

    arXiv 2023

  5. [5]

    Taming transformers for high-resolution image synthesis

    Patrick Esser, Robin Rombach, and Bjorn Ommer. Taming transformers for high-resolution image synthesis. In Pro- ceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 12873–12883, 2021. 3

    2021

  6. [6]

    Vqfr: Blind face restoration with vector-quantized dictionary and parallel de- coder

    Yuchao Gu, Xintao Wang, Liangbin Xie, Chao Dong, Gen Li, Ying Shan, and Ming-Ming Cheng. Vqfr: Blind face restoration with vector-quantized dictionary and parallel de- coder. In ECCV, pages 126–143. Springer, 2022. 3, 7

    work page 2022

  7. [7]

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

    Martin Heusel, Hubert Ramsauer, Thomas Unterthiner, Bernhard Nessler, and Sepp Hochreiter. Gans trained by a two time-scale update rule converge to a local nash equilib- rium. In NeurIPS, 2017. 7

    2017

  8. [8]

    Denoising dif- fusion probabilistic models

    Jonathan Ho, Ajay Jain, and Pieter Abbeel. Denoising dif- fusion probabilistic models. In NeurIPS, pages 6840–6851,

Show all 55 references
  1. [9]

    Labeled faces in the wild: A database for studying face recognition in unconstrained environments

    Gary B Huang, Marwan Mattar, Tamara Berg, and Eric Learned-Miller. Labeled faces in the wild: A database for studying face recognition in unconstrained environments. In Workshop on faces in’Real-Life’Images: detection, align- ment, and recognition, 2008. 6

    2008

  2. [10]

    Autodir: Automatic all-in-one image restoration with latent diffusion

    Yitong Jiang, Zhaoyang Zhang, Tianfan Xue, and Jinwei Gu. Autodir: Automatic all-in-one image restoration with latent diffusion. arXiv preprint arXiv:2310.10123, 2023. 2

    2023 arXiv

  3. [11]

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

    Tero Karras, Timo Aila, Samuli Laine, and Jaakko Lehtinen. Progressive growing of gans for improved quality, stability, and variation. In ICLR, 2018. 6

    2018

  4. [12]

    A style-based generator architecture for generative adversarial networks

    Tero Karras, Samuli Laine, and Timo Aila. A style-based generator architecture for generative adversarial networks. In CVPR, pages 4401–4410, 2019. 6

    2019

  5. [13]

    Denoising diffusion restoration models

    Bahjat Kawar, Michael Elad, Stefano Ermon, and Jiaming Song. Denoising diffusion restoration models. Advances in Neural Information Processing Systems, 35:23593–23606,

  6. [14]

    All-in-one image restoration for unknown corruption

    Boyun Li, Xiao Liu, Peng Hu, Zhongqin Wu, Jiancheng Lv, and Xi Peng. All-in-one image restoration for unknown corruption. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pages 17452– 17462, 2022. 3

    2022

  7. [15]

    Blind face restoration via deep multi-scale component dictionaries

    Xiaoming Li, Chaofeng Chen, Shangchen Zhou, Xianhui Lin, Wangmeng Zuo, and Lei Zhang. Blind face restoration via deep multi-scale component dictionaries. pages 399–415,

  8. [16]

    Learning dual memory dictionaries for blind face restoration

    Xiaoming Li, Shiguang Zhang, Shangchen Zhou, Lei Zhang, and Wangmeng Zuo. Learning dual memory dictionaries for blind face restoration. IEEE Transactions on Pattern Analy- sis and Machine Intelligence, 45(5):5904–5917, 2022. 7

    2022

  9. [17]

    Ucip: A universal frame- work for compressed image super-resolution using dynamic prompt

    Xin Li, Bingchen Li, Yeying Jin, Cuiling Lan, Hanxin Zhu, Yulin Ren, and Zhibo Chen. Ucip: A universal frame- work for compressed image super-resolution using dynamic prompt. arXiv preprint arXiv:2407.13108, 2024. 2

    2024 arXiv

  10. [18]

    Diffbir: Towards blind image restoration with generative diffusion prior

    Xinqi Lin, Jingwen He, Ziyan Chen, Zhaoyang Lyu, Ben Fei, Bo Dai, Wanli Ouyang, Yu Qiao, and Chao Dong. Diffbir: Towards blind image restoration with generative diffusion prior. arXiv preprint arXiv:2308.15070, 2023. 3, 7

    2023 arXiv

  11. [19]

    Preserving full degrada- tion details for blind image super-resolution

    Hongda Liu, Longguang Wang, Ye Zhang, Kaiwen Xue, Shunbo Zhou, and Yulan Guo. Preserving full degrada- tion details for blind image super-resolution. arXiv preprint arXiv:2407.01299, 2024. 3, 4

    2024

  12. [20]

    The dawn of kan in image-to-image (i2i) translation: Integrat- ing kolmogorov-arnold networks with gans for unpaired i2i translation

    Arpan Mahara, Naphtali D Rishe, and Liangdong Deng. The dawn of kan in image-to-image (i2i) translation: Integrat- ing kolmogorov-arnold networks with gans for unpaired i2i translation. arXiv preprint arXiv:2408.08216, 2024. 3

    2024 arXiv

  13. [21]

    Pulse: Self-supervised photo upsam- pling via latent space exploration of generative models

    Sachit Menon, Alexandru Damian, Shijia Hu, Nikhil Ravi, and Cynthia Rudin. Pulse: Self-supervised photo upsam- pling via latent space exploration of generative models. In CVPR, pages 2437–2445, 2020. 3

    2020

  14. [22]

    Conditional generative adversarial nets

    Mehdi Mirza. Conditional generative adversarial nets. arXiv preprint arXiv:1411.1784, 2014. 3

    2014 arXiv

  15. [23]

    completely blind

    Anish Mittal, Rajiv Soundararajan, and Alan C Bovik. Mak- ing a “completely blind” image quality analyzer. IEEE Sig- nal processing letters, 20(3):209–212, 2012. 7

    2012

  16. [24]

    Improved denoising diffusion probabilistic models

    Alexander Quinn Nichol and Prafulla Dhariwal. Improved denoising diffusion probabilistic models. In ICML, pages 8162–8171, 2021. 3

    2021

  17. [25]

    Diffusion autoen- coders: Toward a meaningful and decodable representation

    Konpat Preechakul, Nattanat Chatthee, Suttisak Wizad- wongsa, and Supasorn Suwajanakorn. Diffusion autoen- coders: Toward a meaningful and decodable representation. In Proceedings of the IEEE/CVF conference on computer vi- sion and pattern recognition, pages 10619–10629, 2022. 3

    2022

  18. [26]

    High-resolution image synthesis with latent diffusion models

    Robin Rombach, Andreas Blattmann, Dominik Lorenz, Patrick Esser, and Bj ¨orn Ommer. High-resolution image synthesis with latent diffusion models. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 10684–10695, 2022. 3, 4

    2022

  19. [27]

    Photorealistic text-to-image diffusion models with deep language understanding

    Chitwan Saharia, William Chan, Saurabh Saxena, Lala Li, Jay Whang, Emily L Denton, Kamyar Ghasemipour, Raphael Gontijo Lopes, Burcu Karagol Ayan, Tim Salimans, et al. Photorealistic text-to-image diffusion models with deep language understanding. Advances in neural information...

    2022

  20. [28]

    Image super-resolution via iterative refinement

    Chitwan Saharia, Jonathan Ho, William Chan, Tim Sal- imans, David J Fleet, and Mohammad Norouzi. Image super-resolution via iterative refinement. IEEE Transactions on Pattern Analysis and Machine Intelligence , 45(4):4713– 4726, 2022. 3

    2022

  21. [29]

    Very deep convolutional networks for large-scale image recognition

    Karen Simonyan. Very deep convolutional networks for large-scale image recognition. arXiv preprint arXiv:1409.1556, 2014. 4 10

    2014 arXiv

  22. [30]

    Denoising diffusion implicit models

    Jiaming Song, Chenlin Meng, and Stefano Ermon. Denoising diffusion implicit models. arXiv preprint arXiv:2010.02502, 2020. 3

    2010 arXiv

  23. [31]

    O2v-mapping: Online open-vocabulary mapping with neural implicit representation

    Muer Tie, Julong Wei, Zhengjun Wang, Ke Wu, Shansuai Yuan, Kaizhao Zhang, Jie Jia, Jieru Zhao, Zhongxue Gan, and Wenchao Ding. O2v-mapping: Online open-vocabulary mapping with neural implicit representation. arXiv preprint arXiv:2404.06836, 2024. 3

    2024 arXiv

  24. [32]

    Promptre- storer: A prompting image restoration method with degrada- tion perception

    Cong Wang, Jinshan Pan, Wei Wang, Jiangxin Dong, Mengzhu Wang, Yakun Ju, and Junyang Chen. Promptre- storer: A prompting image restoration method with degrada- tion perception. Advances in Neural Information Processing Systems, 36:8898–8912, 2023. 2

    2023

  25. [33]

    Unsuper- vised degradation representation learning for blind super- resolution

    Longguang Wang, Yingqian Wang, Xiaoyu Dong, Qingyu Xu, Jungang Yang, Wei An, and Yulan Guo. Unsuper- vised degradation representation learning for blind super- resolution. In CVPR, 2021. 2, 3

    2021

  26. [34]

    Promptrr: Diffusion models as prompt generators for single image reflection removal

    Tao Wang, Wanglong Lu, Kaihao Zhang, Wenhan Luo, Tae- Kyun Kim, Tong Lu, Hongdong Li, and Ming-Hsuan Yang. Promptrr: Diffusion models as prompt generators for single image reflection removal. arXiv preprint arXiv:2402.02374,

    arXiv

  27. [35]

    To- wards real-world blind face restoration with generative facial prior

    Xintao Wang, Yu Li, Honglun Zhang, and Ying Shan. To- wards real-world blind face restoration with generative facial prior. pages 9168–9178, 2021. 2, 3, 6, 7

    2021

  28. [36]

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

    Yinhuai Wang, Jiwen Yu, and Jian Zhang. Zero-shot image restoration using denoising diffusion null-space model. In The Eleventh International Conference on Learning Repre- sentations, 2023. 2, 3

    2023

  29. [37]

    Image quality assessment: from error visibility to structural similarity

    Zhou Wang, Alan C Bovik, Hamid R Sheikh, and Eero P Si- moncelli. Image quality assessment: from error visibility to structural similarity. IEEE transactions on image processing, 13(4):600–612, 2004. 7

    2004

  30. [38]

    Restoreformer: High-quality blind face restoration from undegraded key-value pairs

    Zhouxia Wang, Jiawei Zhang, Runjian Chen, Wenping Wang, and Ping Luo. Restoreformer: High-quality blind face restoration from undegraded key-value pairs. InProceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 17512–17521, 2022. 3, 7

    2022

  31. [39]

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

    Zhixin Wang, Ziying Zhang, Xiaoyun Zhang, Huangjie Zheng, Mingyuan Zhou, Ya Zhang, and Yanfeng Wang. Dr2: Diffusion-based robust degradation remover for blind face restoration. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages 1704– 1...

    2023

  32. [40]

    Unsupervised real-world image super resolution via domain-distance aware training

    Yunxuan Wei, Shuhang Gu, Yawei Li, Radu Timofte, Long- cun Jin, and Hengjie Song. Unsupervised real-world image super resolution via domain-distance aware training. In Pro- ceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 13385–13394, 2021. 2, 3

    2021

  33. [41]

    Deflow: Learning complex im- age degradations from unpaired data with conditional flows

    Valentin Wolf, Andreas Lugmayr, Martin Danelljan, Luc Van Gool, and Radu Timofte. Deflow: Learning complex im- age degradations from unpaired data with conditional flows. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 94–103, 2021. 3

    2021

  34. [42]

    Diffir: Efficient diffusion model for image restoration

    Bin Xia, Yulun Zhang, Shiyin Wang, Yitong Wang, Xing- long Wu, Yapeng Tian, Wenming Yang, and Luc Van Gool. Diffir: Efficient diffusion model for image restoration. In Proceedings of the IEEE/CVF International Conference on Computer Vision, pages 13095–13105, 2023. 2, 3

    2023

  35. [43]

    Pgdiff: Guiding diffusion models for versatile face restoration via partial guidance

    Peiqing Yang, Shangchen Zhou, Qingyi Tao, and Chen Change Loy. Pgdiff: Guiding diffusion models for versatile face restoration via partial guidance. Advances in Neural Information Processing Systems, 36, 2024. 3, 7

    2024

  36. [44]

    Wider face: A face detection benchmark

    Shuo Yang, Ping Luo, Chen-Change Loy, and Xiaoou Tang. Wider face: A face detection benchmark. In Proceedings of the IEEE conference on computer vision and pattern recog- nition, pages 5525–5533, 2016. 6

    2016

  37. [45]

    Gan prior embedded network for blind face restoration in the wild

    Tao Yang, Peiran Ren, Xuansong Xie, and Lei Zhang. Gan prior embedded network for blind face restoration in the wild. In CVPR, pages 672–681, 2021. 2, 3, 7

    2021

  38. [46]

    Neural degradation representation learning for all-in-one image restoration

    Mingde Yao, Ruikang Xu, Yuanshen Guan, Jie Huang, and Zhiwei Xiong. Neural degradation representation learning for all-in-one image restoration. arXiv preprint arXiv:2310.12848, 2023. 3

    2023 arXiv

  39. [47]

    Seek for incantations: Towards accurate text-to-image diffusion synthesis through prompt engineer- ing

    Chang Yu, Junran Peng, Xiangyu Zhu, Zhaoxiang Zhang, Qi Tian, and Zhen Lei. Seek for incantations: Towards accurate text-to-image diffusion synthesis through prompt engineer- ing. arXiv preprint arXiv:2401.06345, 2024. 2

    2024 arXiv

  40. [48]

    Difface: Blind face restoration with diffused error contraction

    Zongsheng Yue and Chen Change Loy. Difface: Blind face restoration with diffused error contraction. 2022. 3, 7

    2022

  41. [49]

    Effi- cient diffusion model for image restoration by residual shift- ing

    Zongsheng Yue, Jianyi Wang, and Chen Change Loy. Effi- cient diffusion model for image restoration by residual shift- ing. arXiv preprint arXiv:2403.07319, 2024. 3

    2024 arXiv

  42. [50]

    Efficient degradation-aware any image restoration

    Eduard Zamfir, Zongwei Wu, Nancy Mehta, Danda Dani Paudel, Yulun Zhang, and Radu Timofte. Efficient degradation-aware any image restoration. arXiv preprint arXiv:2405.15475, 2024. 3

    2024 arXiv

  43. [51]

    Styleswin: Transformer-based gan for high-resolution image genera- tion

    Bowen Zhang, Shuyang Gu, Bo Zhang, Jianmin Bao, Dong Chen, Fang Wen, Yong Wang, and Baining Guo. Styleswin: Transformer-based gan for high-resolution image genera- tion. In Proceedings of the IEEE/CVF conference on com- puter vision and pattern recognition , pages 11304–11314,

  44. [52]

    Learning a single convolutional super-resolution network for multiple degradations

    Kai Zhang, Wangmeng Zuo, and Lei Zhang. Learning a single convolutional super-resolution network for multiple degradations. In CVPR, pages 3262–3271, 2018. 6

    2018

  45. [53]

    The unreasonable effectiveness of deep features as a perceptual metric

    Richard Zhang, Phillip Isola, Alexei A Efros, Eli Shechtman, and Oliver Wang. The unreasonable effectiveness of deep features as a perceptual metric. In CVPR, pages 586–595,

  46. [54]

    Selective hourglass mapping for universal image restoration based on diffusion model

    Dian Zheng, Xiao-Ming Wu, Shuzhou Yang, Jian Zhang, Jian-Fang Hu, and Wei-Shi Zheng. Selective hourglass mapping for universal image restoration based on diffusion model. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , pages 25445– 25455, 2024. 3

    2024

  47. [55]

    Towards robust blind face restora- tion with codebook lookup transformer

    Shangchen Zhou, Kelvin Chan, Chongyi Li, and Chen Change Loy. Towards robust blind face restora- tion with codebook lookup transformer. Advances in Neural Information Processing Systems, 35:30599–30611, 2022. 3, 7 11

    2022

Pith tools

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