Pith. sign in

REVIEW 4 major objections 6 minor 5 cited by

DSPO: Direct Semantic Preference Optimization for Real-World Image Super-Resolution

T0 review · 4 major / 6 minor · reviewed 2026-08-16 · 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 applying DPO at the semantic-instance level, with negative text feedback, aligns real-world super-resolution models with human preferences, raising average win rates to 73.6% and 75.6% on DRealSR and RealSR.

desk verdict A legitimate instance-level DPO adaptation for real-world SR with strong human win rates, but the unspecified mask-correspondence step and circular automatic-IQA evaluation leave the central claim only partially verified. read the letter →

arxiv 2504.15176 v1 pith:GTFWAZNE submitted 2025-04-21 cs.CV

Miaomiao Cai , Simiao Li , Wei Li , Xudong Huang , Hanting Chen , Jie Hu , Yunhe Wang This is my paper · ORCID
classification cs.CV
keywords real-worldimagesuper-resolutiondirectpreferenceoptimizationsemanticinstancealignmenthumanhallucinationsuppressiondiffusionmodelspromptinjectioninstance-levelDPO
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

This paper claims that human preference alignment can be brought to real-world image super-resolution by applying Direct Preference Optimization not to whole images but to semantic instances within them. It identifies a core conflict: image-level preference judgments reward outputs that are sharp overall yet hallucinate or artifact in local regions, which fights the pixel-level reconstruction objective. To resolve it, the proposed Direct Semantic Preference Optimization (DSPO) segments each upscaled output into instances, builds preferred and dispreferred pairs instance-by-instance, weights the DPO loss by instance area, and injects user-selected text descriptions of hallucinated instances as negative prompts. The paper reports that this raises human preference win rates over the pre-trained model to 73.6% and 75.6% on DRealSR and RealSR with a one-step SR model, and improves perception and fidelity metrics on both one-step and multi-step frameworks. A reader would care because it offers a way to align generative restoration models with human taste without training a separate reward model.

What carries the argument

The load-bearing object is the mask-weighted, instance-level DPO loss (Eq. 5): for each semantic region $s_m$, the denoising error difference between the current model and the reference model is computed only inside the mask, weighted by $w_m = |s_m| / \sum_m |s_m|$, and passed through a log-sigmoid margin. SAM supplies the masks that define 'the same instance' across outputs; the reference model's per-region denoising error acts as a distribution-drift anchor; the user-description feedback supplies negative prompts constructed from BLIP captions of hallucinated regions. The claim is that this machinery makes the preference signal fine-grained enough to coexist with reconstruction objectives.

What would settle it

Run SAM on the four SR outputs for a set of LQ images, align the outputs, and compute the overlap between the mask used for the preferred image and the mask used for the dispreferred image; if those masks do not correspond to the same scene instance, the instance-level preference pairs are comparing different content and the reported win-rate gains should be re-examined.

Watch

Extended reading notes

Core claim

On its own terms, the paper's central claim is that the mismatch between holistic human preferences and pixel-level super-resolution objectives can be dissolved by making preferences instance-level and semantic. DSPO generates several SR outputs per low-quality input, segments each with SAM, selects best and worst outputs within each instance region under human or automatic evaluation, and optimizes the diffusion denoising objective with a DPO-style margin weighted by the relative area of each semantic instance. The same pipeline adds a hallucination-suppression branch: a vision-language model describes each instance, and descriptions of instances judged inconsistent with the input are stored as negative text feedback and injected into the prompt branch during training. The paper's empirical assertion is that this combination, applied as a plug-in to a one-step diffusion SR model and a multi-step text-to-image backbone, outperforms the pre-trained models, SFT, DDPO, and Diffusion-DPO on human preference win rates and on the reported IQA metrics.

Load-bearing premise

The load-bearing premise is that the SAM mask of an instance in one SR output and the SAM mask of the 'same' instance in another output refer to the same real-world object; if the masks are never registered or matched across outputs, the preferred and dispreferred pair compares different content and the mask-weighted DPO loss is computed on mislabeled regions.

Editorial extensions

If this is right

  • Instance-level semantic preference alignment can be added to existing one-step and multi-step diffusion SR models without changing inference-time cost, since the alignment is a training-phase objective.
  • Human annotators prefer DSPO outputs over the pre-trained model roughly three-quarters of the time on real-world benchmarks, and over SFT, DDPO, and Diffusion-DPO as well.
  • Automatic quality metrics improve alongside human preference, so the method does not merely chase annotator taste.
  • The negative text-feedback branch contributes separately to artifact suppression, since removing it degrades performance in the ablation.

Reading between the lines

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

  • The paper never specifies how SAM masks are registered across different SR outputs; a testable extension would check mask correspondence, since the instance-level pairs are the whole basis of the loss.
  • A practical extension would replace human annotation with learned IQA scores entirely and measure whether the win-rate gains survive; the paper already uses an automatic IQA variant, so the human step may be compressible.
  • The same instance-level preference recipe could transfer to other dense generative restoration tasks where local hallucinations dominate, such as deblurring or deraining.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

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 manuscript introduces DSPO, a preference-alignment method for real-world image super-resolution that applies Direct Preference Optimization at the semantic-instance level rather than the whole-image level. Given a low-quality input, the pre-trained SR model generates several candidate outputs; each output is segmented with SAM into instance regions, and per-instance preferred/dispreferred pairs are selected either by human annotators or by an automatic sum of eight IQA metrics. The training loss is a masked diffusion-DPO objective (Eq. 5) with instance-size weights (Eq. 3), and a user-description feedback strategy is proposed to suppress hallucinations. Experiments compare DSPO against SFT, DDPO, and Diffusion-DPO on one-step OSEDiff and multi-step SD2-base frameworks; human win rates of 73.6% on DRealSR and 75.6% on RealSR are reported for the one-step framework.

Significance. The contribution is potentially valuable: it is a plausible first application of DPO-style preference alignment to real-world image SR, and the idea of per-instance preferred/dispreferred preferences over whole-image preferences addresses a genuine tension between pixel-level reconstruction and image-level reward. The human study is an independent, meaningful signal, and reporting 95% confidence intervals is a strength. The paper also provides ablations of the two proposed strategies. However, the current manuscript does not specify how SAM masks are matched across candidate images, and the automatic-IQA evaluation is circular with respect to the preference-label construction. These issues must be resolved before the central claims can be accepted.

major comments (4)
  1. [Sec. 4.2.1 (mask correspondence)] The construction of instance-level preference pairs assumes that the same semantic region s_m can be compared across the N SR outputs, but no matching or registration procedure is specified. SAM applied independently to each SR output yields an unordered, image-dependent set of masks: mask indices can differ, segment counts can differ, and hallucinated content can introduce spurious regions. If the 'same s_m' across outputs is selected by mask index only, the preferred/dispreferred pairs in Eq. (4) and the masked losses in Eq. (5) can compare non-corresponding physical regions. Because all 2,500 annotated instance regions and the reported human win rates depend on this correspondence, the paper must define the matching rule (e.g., IoU-based mask matching, a shared mask from the LQ/HQ image, or post-hoc association), and must state that the same rule was used in both the human and automatic annotation pipelines.
  2. [Sec. 5.1.1 vs. Sec. 5.2.2] The automatic preference labels are created by summing the same eight normalized IQA metrics (PSNR, SSIM, LPIPS, DISTS, NIQE, MUSIQ, MANIQA, CLIPIQA) that are then used as the evaluation metrics in Sec. 5.2.2. Training on labels derived from these metrics can only increase the same metrics, so the reported 'improvements across all metrics' are partly self-fulfilling and do not constitute independent evidence of quality. The human win rates in Sec. 5.2.1 are independent and should remain the primary evidence; the automatic-IQA comparison should either use held-out metrics not involved in label construction or be explicitly framed as a training-signal consistency check.
  3. [Sec. 4.3 and Eq. (8)] The user-description feedback strategy is not defined at the level of the optimization objective. Eq. (5) contains no term involving the negative text prompt, while Eq. (8) only writes L(x_w, x_l | x_LQ, s_m, p_negative) without specifying how p_negative conditions the diffusion model or enters the preference loss. If the feedback prompts are used only at inference time, this should be stated and the inference-time mechanism described; if they are used during training, the exact loss term must be given. Without this, the reported hallucination suppression from the M2 ablation is not attributable to a concrete mechanism.
  4. [Sec. 4.2.1, Eq. (4)] The model probability p_theta(x | x_LQ, s_m) is not defined, and it is unclear whether the semantic mask is a conditioning input to the diffusion model. The architecture description and inference procedure mention only the LQ image and cfg scale; Eq. (5) applies the mask as a spatial weight on the denoising error, not as a conditioning signal. Please state explicitly whether p_theta is conditioned on s_m, or reformulate Eq. (4) as a masked weighted DPO loss and clarify how the DPO likelihood ratio is computed in that case.
minor comments (6)
  1. [Abstract and Sec. 1] 'may leading to' should be 'may lead to'.
  2. [Eq. (2)] The sentence 'Here, condition text is compactness.' appears garbled and should be removed or rewritten; the condition c is not defined in the diffusion-DPO context.
  3. [Sec. 2.3] 'DDPO [9]' is cited to Reference [9], which is the DDPM paper (Ho et al.), not the DDPO paper; the reference list needs correction.
  4. [Sec. 4.2.1] The notation sum_{m=1}^M s_m = 1 implies the SAM masks form a partition, but SAM masks typically overlap and may omit background; please specify any post-processing used to make the masks disjoint.
  5. [Sec. 5.1.1] The statement 'resulting in a total of 2500 instance regions for annotation, as each LQ has four corresponding SR results' is ambiguous; please clarify whether 2,500 refers to instances (500 images x 5 instances) or to instance-image pairs including the four candidates.
  6. [Table 1 caption] 'different different optimization objectives' contains a duplicated word.

Circularity Check

1 steps flagged · score 6.0 of 10

Automatic-IQA evaluation reuses the exact metric sum that generated the training preference labels, making those metric gains partly self-fulfilling; the human win-rate claim remains independent.

  1. fitted input called prediction [Sec. 5.1.1 (Automatic IQA Method) and Sec. 5.2.2 (Preference Alignment from Automatic IQA)]
    "We first normalize the eight metrics individually (e.g. PSNR, SSIM, LPIPS, DISTS, NIQE, MUSIQ, MANIQA, and CLIPIQA)... Then all metrics are adjusted to be a positive trend and summed to obtain an overall score, simulating human preference ratings. ... We utilize a series of metrics to evaluate the results of automatic IQA method and other method, including PSNR, SSIM, LPIPS, DISTS, NIQE, MUSIQ, MANIQA, and CLIPIQA."

    The training preference labels for the automatic branch are not independent of the evaluation: Sec. 5.1.1 defines the preferred/dispreferred instances by summing normalized values of exactly the eight metrics, then Sec. 5.2.2 reports improvements on those same eight metrics as evidence. Because the model is trained to increase the probability of images that score higher on this sum, the reported per-metric gains are partially enforced by the label construction rather than being out-of-sample predictions. The human-annotator win rates in Sec. 5.2.1 are independent and remain valid; the automatic-IQA section, however, evaluates with the training objective itself.

full rationale

The only load-bearing circular step is the automatic-IQA evaluation reusing the metric sum that generated its labels. The paper's core novelty and headline numbers rest on human annotator preference win rates (73.6% and 75.6%), which use fresh comparisons on DRealSR and RealSR and do not reduce to the training labels. The instance-mask correspondence issue noted in Sec. 4.2.1 is a missing-specification and correctness concern, not a circularity: no equation reduces to an input. No self-citation chain is load-bearing. Therefore the partial circularity score is 6: the automatic-IQA evidence is self-fulfilling, while the human evidence is independent and keeps the central claim from being entirely circular.

Assumptions & free parameters 4 free parameters · 6 assumptions · 0 invented entities

The training signal depends on hand-set constants (beta, N=4, top-5, threshold 0.1) and on unverified modeling assumptions about SAM mask correspondence, BLIP text feedback, and the validity of the composite IQA score as a human-preference proxy. No invented physical entities are introduced.

free parameters (4)
  • DPO temperature beta = 8000
    Set to 8000 in Sec. 5.1.1; controls preference strength and is chosen by hand.
  • Candidate outputs per LQ (N) = 4
    Four SR results generated per LQ; determined by 'a series of experiments' in Sec. 5.1.1.
  • Annotated instances per image = 5
    Top 5 largest SAM segments per image are annotated (Sec. 5.1.1).
  • BLIP similarity hallucination threshold = 0.1
    Text similarity below 0.1 marks an instance as hallucinated (Sec. 5.1.1).
assumptions (6)
  • domain assumption SAM segmentation produces consistent, corresponding instance masks across different SR outputs of the same LQ image.
    Sec. 4.2.1 compares 'the same instance sm region across different images' but does not specify or validate mask registration across outputs.
  • domain assumption Diffusion-DPO's noise-prediction loss remains a valid preference objective when masked per semantic instance.
    Sec. 4.2.2 adapts Eq. 2 to Eq. 5 via Ldiff terms; this transfer is asserted, not derived.
  • domain assumption Four generated candidates per LQ are sufficient for Best/Worst-of-N to reveal stable instance-level preferences.
    Sec. 5.1.1 fixes N=4 from 'a series of experiments'; no sampling analysis is provided.
  • domain assumption BLIP descriptions can detect hallucinated content, and injecting negative text prompts steers the SR model away from it.
    Sec. 4.3 and Sec. 5.1.1 use BLIP similarity threshold 0.1; the prompt injection mechanism is not concretely specified.
  • ad hoc to paper The equal-weight sum of eight normalized IQA metrics simulates human preference for automatic label generation.
    Sec. 5.1.1 builds automatic preferences from these metrics; the composite score is not validated as a preference proxy.
  • domain assumption Using the reference model's noise-prediction error as a regularizer prevents distribution drift during DSPO training.
    Stated at Eq. 5 and 6; no stability analysis is given.

how reviews work

0 comments
Cite this review

Pith. "Pith review of DSPO: Direct Semantic Preference Optimization for Real-World Image Super-Resolution." pith.science (2026). https://pith.science/paper/GTFWAZNE

@misc{pith2026250415176,
  author       = {Pith},
  title        = {Pith review of: DSPO: Direct Semantic Preference Optimization for Real-World Image Super-Resolution},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/GTFWAZNE}},
  note         = {Machine review of arXiv:2504.15176}
}
read the original abstract

Recent advances in diffusion models have improved Real-World Image Super-Resolution (Real-ISR), but existing methods lack human feedback integration, risking misalignment with human preference and may leading to artifacts, hallucinations and harmful content generation. To this end, we are the first to introduce human preference alignment into Real-ISR, a technique that has been successfully applied in Large Language Models and Text-to-Image tasks to effectively enhance the alignment of generated outputs with human preferences. Specifically, we introduce Direct Preference Optimization (DPO) into Real-ISR to achieve alignment, where DPO serves as a general alignment technique that directly learns from the human preference dataset. Nevertheless, unlike high-level tasks, the pixel-level reconstruction objectives of Real-ISR are difficult to reconcile with the image-level preferences of DPO, which can lead to the DPO being overly sensitive to local anomalies, leading to reduced generation quality. To resolve this dichotomy, we propose Direct Semantic Preference Optimization (DSPO) to align instance-level human preferences by incorporating semantic guidance, which is through two strategies: (a) semantic instance alignment strategy, implementing instance-level alignment to ensure fine-grained perceptual consistency, and (b) user description feedback strategy, mitigating hallucinations through semantic textual feedback on instance-level images. As a plug-and-play solution, DSPO proves highly effective in both one-step and multi-step SR frameworks.

Figures

Figures reproduced from arXiv: 2504.15176 by the authors.

Figure 1
Figure 1. The dilemma between image-level preferences in DPO [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. The overview of proposed direct semantic preference optimization (DSPO) method. [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. The user preference win rates of DSPO, compared to [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Quantitative comparison of DSPO with baselines on the automatic IQA Method. (a)-(c) depict the radar plots for the one-step [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: Visualization comparison of DSPO and other baseline methods on the human annotator method (Top), and the automatic IQA [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 6
Figure 6. Figure 6: The visualization of DSPO based on perception metrics [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 5 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Zoom-IQA: Image Quality Assessment with Reliable Region-Aware Reasoning

    cs.CV 2026-01 conditional novelty 7.0 of 10

    Zoom-IQA lets a vision-language model iteratively crop and zoom into image regions before giving a quality score, improving reasoning and restoration guidance over single-pass IQA models.

  2. Selective Timestep Weighting and Advantage-Based Replay for Sample-Efficient Diffusion RLHF

    cs.LG 2026-07 conditional novelty 5.0 of 10

    Two plug-and-play strategies — per-timestep advantage weighting and advantage-based trajectory replay — improve diffusion RLHF sample efficiency up to 6× across five reward functions.

  3. Bridging Information Asymmetry: A Hierarchical Framework for Deterministic Blind Face Restoration

    cs.CV 2026-01 conditional novelty 5.0 of 10

    Pref-Restore combines AR semantic tokens, a diffusion generator, and DiffusionNFT-style RL to make blind face restoration more consistent, but its deterministic-identity claim is weakened by self-referential rewards a...

  4. From Black Box to Transparency: Enhancing Automated Interpreting Assessment with Explainable AI in College Classrooms

    cs.CL 2025-08 unverdicted novelty 5.0 of 10

    An explainable model using BLEURT, CometKiwi, pause features, and Chinese phraseological diversity predicts human-rated quality dimensions in English-Chinese consecutive interpreting, with SHAP identifying the stronge...

  5. Hierarchical Fine-grained Preference Optimization for Physically Plausible Video Generation

    cs.CV 2025-08 conditional novelty 4.0 of 10

    A hierarchical direct preference optimization with four alignment levels plus automated data selection improves physical plausibility of text-to-video models.

Reference graph

Works this paper leans on

42 extracted references · 18 canonical work pages · cited by 5 Pith papers

  1. [1]

    Gpt-4 technical report

    Josh Achiam, Steven Adler, Sandhini Agarwal, Lama Ah- mad, Ilge Akkaya, Florencia Leoni Aleman, Diogo Almeida, Janko Altenschmidt, Sam Altman, Shyamal Anadkat, et al. Gpt-4 technical report. arXiv preprint arXiv:2303.08774 ,

    arXiv

  2. [2]

    Ntire 2017 challenge on single image super-resolution: Dataset and study

    Eirikur Agustsson and Radu Timofte. Ntire 2017 challenge on single image super-resolution: Dataset and study. In Pro- ceedings of the IEEE conference on computer vision and pat- tern recognition workshops, pages 126–135, 2017. 6

    2017

  3. [3]

    Training a helpful and harmless assistant with reinforcement learning from human feedback

    Yuntao Bai, Andy Jones, Kamal Ndousse, Amanda Askell, Anna Chen, Nova DasSarma, Dawn Drain, Stanislav Fort, Deep Ganguli, Tom Henighan, et al. Training a helpful and harmless assistant with reinforcement learning from human feedback. arXiv preprint arXiv:2204.05862, 2022. 2

    arXiv 2022

  4. [4]

    Toward real-world single image super-resolution: A new benchmark and a new model

    Jianrui Cai, Hui Zeng, Hongwei Yong, Zisheng Cao, and Lei Zhang. Toward real-world single image super-resolution: A new benchmark and a new model. In Proceedings of the IEEE/CVF international conference on computer vision, pages 3086–3095, 2019. 6

    2019

  5. [5]

    Real-world single image super-resolution: A brief review

    Honggang Chen, Xiaohai He, Linbo Qing, Yuanyuan Wu, Chao Ren, Ray E Sheriff, and Ce Zhu. Real-world single image super-resolution: A brief review. Information Fusion, 79:124–145, 2022. 1

    work page 2022

  6. [6]

    Diffusion models beat gans on image synthesis

    Prafulla Dhariwal and Alexander Nichol. Diffusion models beat gans on image synthesis. Advances in neural informa- tion processing systems, 34:8780–8794, 2021. 1

    2021

  7. [7]

    Image quality assessment: Unifying structure and texture similarity

    Keyan Ding, Kede Ma, Shiqi Wang, and Eero P Simoncelli. Image quality assessment: Unifying structure and texture similarity. IEEE transactions on pattern analysis and ma- chine intelligence, 44(5):2567–2581, 2020. 5, 6

    work page 2020

  8. [8]

    Learning a deep convolutional network for image super-resolution

    Chao Dong, Chen Change Loy, Kaiming He, and Xiaoou Tang. Learning a deep convolutional network for image super-resolution. In Computer Vision–ECCV 2014: 13th European Conference, Zurich, Switzerland, September 6-12, 2014, Proceedings, Part IV 13 , pages 184–199. Springer,

    2014

Show all 42 references
  1. [9]

    Denoising dif- fusion probabilistic models

    Jonathan Ho, Ajay Jain, and Pieter Abbeel. Denoising dif- fusion probabilistic models. Advances in neural information processing systems, 33:6840–6851, 2020. 1, 2, 4, 6

    2020

  2. [10]

    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,

  3. [11]

    Musiq: Multi-scale image quality transformer

    Junjie Ke, Qifei Wang, Yilin Wang, Peyman Milanfar, and Feng Yang. Musiq: Multi-scale image quality transformer. In Proceedings of the IEEE/CVF international conference on computer vision, pages 5148–5157, 2021. 5, 6

    2021

  4. [12]

    Variational diffusion models

    Diederik Kingma, Tim Salimans, Ben Poole, and Jonathan Ho. Variational diffusion models. Advances in neural infor- mation processing systems, 34:21696–21707, 2021. 4

    2021

  5. [13]

    Segment any- thing

    Alexander Kirillov, Eric Mintun, Nikhila Ravi, Hanzi Mao, Chloe Rolland, Laura Gustafson, Tete Xiao, Spencer White- head, Alexander C Berg, Wan-Yen Lo, et al. Segment any- thing. In Proceedings of the IEEE/CVF international con- ference on computer vision , pages 4015–4026, 20...

    2023

  6. [14]

    Photo- realistic single image super-resolution using a generative ad- versarial network

    Christian Ledig, Lucas Theis, Ferenc Husz´ar, Jose Caballero, Andrew Cunningham, Alejandro Acosta, Andrew Aitken, Alykhan Tejani, Johannes Totz, Zehan Wang, et al. Photo- realistic single image super-resolution using a generative ad- versarial network. In Proceedings of the IE...

  7. [15]

    Blip: Bootstrapping language-image pre-training for unified vision-language understanding and generation

    Junnan Li, Dongxu Li, Caiming Xiong, and Steven Hoi. Blip: Bootstrapping language-image pre-training for unified vision-language understanding and generation. In Interna- tional conference on machine learning, pages 12888–12900. PMLR, 2022. 5

    2022

  8. [16]

    Aligning diffusion mod- els by optimizing human utility

    Shufan Li, Konstantinos Kallidromitis, Akash Gokul, Yusuke Kato, and Kazuki Kozuka. Aligning diffusion mod- els by optimizing human utility. Advances in Neural Infor- mation Processing Systems, 37:24897–24925, 2025. 1

    2025

  9. [17]

    Lsdir: A large scale dataset for image restoration

    Yawei Li, Kai Zhang, Jingyun Liang, Jiezhang Cao, Ce Liu, Rui Gong, Yulun Zhang, Hao Tang, Yun Liu, Denis Deman- dolx, et al. Lsdir: A large scale dataset for image restoration. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 1775–17...

    2023

  10. [18]

    Diff- bir: Toward blind image restoration with generative diffusion prior

    Xinqi Lin, Jingwen He, Ziyan Chen, Zhaoyang Lyu, Bo Dai, Fanghua Yu, Yu Qiao, Wanli Ouyang, and Chao Dong. Diff- bir: Toward blind image restoration with generative diffusion prior. In European Conference on Computer Vision , pages 430–448. Springer, 2024. 1

    2024

  11. [19]

    Decoupled weight decay regularization

    Ilya Loshchilov and Frank Hutter. Decoupled weight decay regularization. arXiv preprint arXiv:1711.05101, 2017. 5

    2017 arXiv

  12. [20]

    Improved denoising diffusion probabilistic models

    Alexander Quinn Nichol and Prafulla Dhariwal. Improved denoising diffusion probabilistic models. In International conference on machine learning, pages 8162–8171. PMLR,

  13. [21]

    Training language models to follow instructions with human feedback

    Long Ouyang, Jeffrey Wu, Xu Jiang, Diogo Almeida, Car- roll Wainwright, Pamela Mishkin, Chong Zhang, Sandhini Agarwal, Katarina Slama, Alex Ray, et al. Training language models to follow instructions with human feedback. Ad- vances in neural information processing systems, 35:...

    2022

  14. [22]

    Direct preference optimization: Your language model is secretly a reward model

    Rafael Rafailov, Archit Sharma, Eric Mitchell, Christo- pher D Manning, Stefano Ermon, and Chelsea Finn. Direct preference optimization: Your language model is secretly a reward model. Advances in Neural Information Processing Systems, 36:53728–53741, 2023. 2, 3

    2023

  15. [23]

    Image super- resolution via iterative refinement

    Chitwan Saharia, Jonathan Ho, William Chan, Tim Sali- mans, David J Fleet, and Mohammad Norouzi. Image super- resolution via iterative refinement. IEEE transactions on pattern analysis and machine intelligence, 45(4):4713–4726,

  16. [24]

    Proximal policy optimization algo- rithms

    John Schulman, Filip Wolski, Prafulla Dhariwal, Alec Rad- ford, and Oleg Klimov. Proximal policy optimization algo- rithms. arXiv preprint arXiv:1707.06347, 2017. 2

    2017 arXiv

  17. [25]

    Generative modeling by esti- mating gradients of the data distribution

    Yang Song and Stefano Ermon. Generative modeling by esti- mating gradients of the data distribution. Advances in neural information processing systems, 32, 2019. 4

    2019

  18. [26]

    Score-based generative modeling through stochastic differential equa- tions

    Yang Song, Jascha Sohl-Dickstein, Diederik P Kingma, Ab- hishek Kumar, Stefano Ermon, and Ben Poole. Score-based generative modeling through stochastic differential equa- tions. arXiv preprint arXiv:2011.13456, 2020. 1, 2

    2011 arXiv

  19. [28]

    Llama 2: Open foundation and fine-tuned chat models.arXiv preprint arXiv:2307.09288, 2023

    Hugo Touvron, Louis Martin, Kevin Stone, Peter Albert, Amjad Almahairi, Yasmine Babaei, Nikolay Bashlykov, Soumya Batra, Prajjwal Bhargava, Shruti Bhosale, et al. Llama 2: Open foundation and fine-tuned chat models.arXiv preprint arXiv:2307.09288, 2023. 1

    2023 arXiv

  20. [29]

    Diffusion model align- ment using direct preference optimization

    Bram Wallace, Meihua Dang, Rafael Rafailov, Linqi Zhou, Aaron Lou, Senthil Purushwalkam, Stefano Ermon, Caiming Xiong, Shafiq Joty, and Nikhil Naik. Diffusion model align- ment using direct preference optimization. In Proceedings of the IEEE/CVF Conference on Computer Vision a...

    2024

  21. [30]

    Ex- ploring clip for assessing the look and feel of images

    Jianyi Wang, Kelvin CK Chan, and Chen Change Loy. Ex- ploring clip for assessing the look and feel of images. InPro- ceedings of the AAAI conference on artificial intelligence , pages 2555–2563, 2023. 5, 6

    2023

  22. [31]

    Exploiting diffusion prior for real-world image super-resolution

    Jianyi Wang, Zongsheng Yue, Shangchen Zhou, Kelvin CK Chan, and Chen Change Loy. Exploiting diffusion prior for real-world image super-resolution. International Journal of Computer Vision, 132(12):5929–5949, 2024. 1, 2

    2024

  23. [32]

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

    Xintao Wang, Liangbin Xie, Chao Dong, and Ying Shan. Real-esrgan: Training real-world blind super-resolution with pure synthetic data. In Proceedings of the IEEE/CVF inter- national conference on computer vision , pages 1905–1914,

    1905

  24. [33]

    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. 5, 6

    2004

  25. [34]

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

    Pengxu Wei, Ziwei Xie, Hannan Lu, Zongyuan Zhan, Qixi- ang Ye, Wangmeng Zuo, and Liang Lin. Component divide- and-conquer for real-world image super-resolution. In Com- puter Vision–ECCV 2020: 16th European Conference, Glas- gow, UK, August 23–28, 2020, Proceedings, Part VIII ...

    2020

  26. [35]

    Seesr: Towards semantics- aware real-world image super-resolution

    Rongyuan Wu, Tao Yang, Lingchen Sun, Zhengqiang Zhang, Shuai Li, and Lei Zhang. Seesr: Towards semantics- aware real-world image super-resolution. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 25456–25467, 2024. 1, 2

    2024

  27. [36]

    One-step effective diffusion network for real-world image super-resolution

    Rongyuan Wu, Lingchen Sun, Zhiyuan Ma, and Lei Zhang. One-step effective diffusion network for real-world image super-resolution. Advances in Neural Information Process- ing Systems, 37:92529–92553, 2025. 1, 2, 5

    2025

  28. [37]

    Imagere- ward: Learning and evaluating human preferences for text- to-image generation

    Jiazheng Xu, Xiao Liu, Yuchen Wu, Yuxuan Tong, Qinkai Li, Ming Ding, Jie Tang, and Yuxiao Dong. Imagere- ward: Learning and evaluating human preferences for text- to-image generation. Advances in Neural Information Pro- cessing Systems, 36:15903–15935, 2023. 2

    2023

  29. [38]

    Maniqa: Multi-dimension attention network for no-reference image quality assessment

    Sidi Yang, Tianhe Wu, Shuwei Shi, Shanshan Lao, Yuan Gong, Mingdeng Cao, Jiahao Wang, and Yujiu Yang. Maniqa: Multi-dimension attention network for no-reference image quality assessment. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pag...

    2022

  30. [39]

    Pixel-aware stable diffusion for realistic image super-resolution and personalized stylization

    Tao Yang, Rongyuan Wu, Peiran Ren, Xuansong Xie, and Lei Zhang. Pixel-aware stable diffusion for realistic image super-resolution and personalized stylization. In European Conference on Computer Vision , pages 74–91. Springer,

  31. [40]

    Scaling up to excellence: Practicing model scaling for photo- realistic image restoration in the wild

    Fanghua Yu, Jinjin Gu, Zheyuan Li, Jinfan Hu, Xiangtao Kong, Xintao Wang, Jingwen He, Yu Qiao, and Chao Dong. Scaling up to excellence: Practicing model scaling for photo- realistic image restoration in the wild. In Proceedings of the IEEE/CVF Conference on Computer Vision and...

    2024

  32. [41]

    A feature-enriched completely blind image quality evaluator.IEEE Transactions on Image Processing, 24(8):2579–2591, 2015

    Lin Zhang, Lei Zhang, and Alan C Bovik. A feature-enriched completely blind image quality evaluator.IEEE Transactions on Image Processing, 24(8):2579–2591, 2015. 5, 6

    2015

  33. [42]

    The unreasonable effectiveness of deep features as a perceptual metric

    Richard Zhang, Phillip Isola, Alexei A Efros, Eli Shecht- man, and Oliver Wang. The unreasonable effectiveness of deep features as a perceptual metric. In Proceedings of the IEEE conference on computer vision and pattern recogni- tion, pages 586–595, 2018. 5, 6

    2018

  34. [43]

    Real-world image super-resolution as multi-task learning

    Wenlong Zhang, Xiaohui Li, Guangyuan Shi, Xiangyu Chen, Yu Qiao, Xiaoyun Zhang, Xiao-Ming Wu, and Chao Dong. Real-world image super-resolution as multi-task learning. Advances in Neural Information Processing Systems , 36: 21003–21022, 2023. 1

    2023

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.