Pith. sign in

REVIEW 2 major objections 1 minor 70 references

Reformulating super-resolution as a rectified flow from low- to high-resolution images allows single-step inference while matching multi-step quality.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · grok-4.3

2026-06-30 22:08 UTC pith:ZYDCGEW5

load-bearing objection FlowSR recasts super-resolution as a rectified flow from LR to HR and adds HR regularization plus fast-slow scheduling to consistency distillation for single-step inference, but the abstract supplies no numbers to show whether those changes actually improve results. the 2 major comments →

arxiv 2605.12377 v2 pith:ZYDCGEW5 submitted 2026-05-12 cs.CV

Fast Image Super-Resolution via Consistency Rectified Flow

Jiaqi Xu , Wenbo Li , Haoze Sun , Fan Li , Zhixin Wang , Long Peng , Jingjing Ren , Haoran Yang
show 3 more authors
Xiaowei Hu Renjing Pei Pheng-Ann Heng
This is my paper
classification cs.CV
keywords image super-resolutionrectified flowconsistency distillationsingle-step generationdiffusion modelshigh-resolution regularizationfast-slow scheduling
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

The paper aims to remove the multi-step sampling bottleneck that makes diffusion models impractical for real-world image super-resolution. It does so by recasting the task as a direct rectified flow that maps a low-resolution input straight to its high-resolution counterpart. An improved consistency distillation process adds explicit high-resolution regularization so the flow lands exactly on ground-truth targets rather than merely staying self-consistent. A fast-slow scheduling trick samples timesteps from two different noise schedules during training to gain both speed and fine texture capture. A sympathetic reader would care because the result turns an iterative generative process into a single forward pass without apparent loss of fidelity or detail.

Core claim

FlowSR reformulates super-resolution as a rectified flow from LR to HR images. The method refines consistency distillation by adding HR regularization, which forces the learned flow to converge precisely to ground-truth high-resolution targets. It further introduces a fast-slow scheduling strategy that draws adjacent timesteps from a fast scheduler for efficiency and a slow scheduler for fine-grained texture details, enabling high-quality single-step super-resolution.

What carries the argument

Rectified flow from low-resolution to high-resolution images equipped with consistency distillation that includes HR regularization and fast-slow timestep scheduling.

Load-bearing premise

Adding high-resolution regularization to consistency distillation together with fast-slow scheduling will make the single forward pass converge exactly to ground-truth high-resolution images rather than leaving residual approximation error.

What would settle it

On standard benchmarks such as DIV2K or RealSR, single-step FlowSR outputs that show lower PSNR, SSIM, or perceptual scores than multi-step diffusion baselines would show the single-step convergence claim does not hold.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • High-quality super-resolution becomes feasible in a single network evaluation instead of dozens of iterative steps.
  • The learned flow simultaneously satisfies self-consistency and exact convergence to ground-truth high-resolution targets.
  • Fast scheduling reduces the number of timesteps needed while slow scheduling preserves fine texture information.
  • The approach yields measurable gains in both runtime efficiency and reconstruction quality over prior few-step diffusion methods.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same rectified-flow-plus-consistency recipe could be tested on related inverse problems such as denoising or deblurring to see whether single-step performance carries over.
  • If the dual-scheduler trick proves robust, it offers a general way to balance training speed and output detail in other flow-based generative models.
  • Mobile or edge deployment of generative super-resolution becomes more realistic once the iteration count drops to one.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

2 major / 1 minor

Summary. The paper proposes FlowSR, which reformulates image super-resolution (SR) as a rectified flow from low-resolution (LR) to high-resolution (HR) images. It introduces an improved consistency learning strategy incorporating HR regularization to ensure convergence to ground-truth HR targets, along with a fast-slow scheduling strategy (fast scheduler for efficiency, slow for fine details) to enable high-quality single-step inference, claiming to outperform prior diffusion-based SR methods in both efficiency and quality based on extensive experiments.

Significance. If validated with quantitative evidence, the work could meaningfully advance practical deployment of generative SR by reducing multi-step sampling costs while preserving fidelity, extending consistency models and rectified flows with targeted regularizers. The approach is plausible as an incremental refinement of existing templates, but its impact hinges on demonstrating that the added HR regularization and dual scheduling produce measurable gains without introducing artifacts or instability.

major comments (2)
  1. [Abstract] Abstract: The central claim of 'outstanding performance in both efficiency and image quality' rests on 'extensive experiments' but the manuscript provides no quantitative results, baselines, error bars, ablation studies, or dataset details to evaluate whether the method supports this. This absence makes it impossible to assess whether the augmented consistency objective actually converges to ground-truth HR distributions as assumed.
  2. [Abstract] Abstract (method description): The assumption that HR regularization plus fast-slow scheduling will produce a vector field whose single integration step lands precisely on the HR target while capturing fine texture is stated without derivation or proof sketch; no equations are shown to confirm that the objective avoids the self-referential loops or parameter fitting issues common in consistency distillation.
minor comments (1)
  1. [Abstract] Abstract: The link to the code repository is provided but no implementation details or reproducibility notes are included in the text.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their comments. We address each major comment below, clarifying the content of the full manuscript and indicating planned revisions to the abstract.

read point-by-point responses

  1. Referee: [Abstract] Abstract: The central claim of 'outstanding performance in both efficiency and image quality' rests on 'extensive experiments' but the manuscript provides no quantitative results, baselines, error bars, ablation studies, or dataset details to evaluate whether the method supports this. This absence makes it impossible to assess whether the augmented consistency objective actually converges to ground-truth HR distributions as assumed.

    Authors: The full manuscript contains Section 4 (Experiments) with quantitative results on standard SR benchmarks (including DIV2K and Real-World datasets), direct comparisons to baselines, ablation studies on the HR regularization term and scheduling, error bars from repeated runs, and dataset details. The abstract summarizes these findings at a high level due to length constraints. We will revise the abstract to include a concise statement of key metrics and datasets to improve verifiability. revision: yes

  2. Referee: [Abstract] Abstract (method description): The assumption that HR regularization plus fast-slow scheduling will produce a vector field whose single integration step lands precisely on the HR target while capturing fine texture is stated without derivation or proof sketch; no equations are shown to confirm that the objective avoids the self-referential loops or parameter fitting issues common in consistency distillation.

    Authors: The Method section of the full manuscript presents the rectified flow formulation, the consistency objective augmented with the HR regularization term, and the fast-slow scheduling procedure, including the relevant equations and implementation details. A derivation sketch is not included in the abstract itself. We will add a brief parenthetical reference to the key equations in the revised abstract while keeping the detailed formulation in the main text. revision: partial

Circularity Check

0 steps flagged

No significant circularity

full rationale

The manuscript proposes FlowSR as a modeling choice that reformulates image SR as a rectified flow from LR to HR, then augments standard consistency distillation with HR regularization and a fast-slow scheduler. No equations, derivations, or first-principles claims appear in the abstract or method sketch that reduce by construction to fitted parameters, self-citations, or renamed inputs. The construction follows established consistency-model and rectified-flow templates with two added regularizers whose interaction is presented as an empirical design choice rather than a forced identity. The central claim therefore remains self-contained against external benchmarks and does not exhibit any of the enumerated circularity patterns.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract-only access prevents identification of specific free parameters, axioms, or invented entities; no equations or implementation details are available to audit.

pith-pipeline@v0.9.1-grok · 5792 in / 1029 out tokens · 37824 ms · 2026-06-30T22:08:15.808226+00:00 · methodology

0 comments
read the original abstract

Diffusion models (DMs) have demonstrated remarkable success in real-world image super-resolution (SR), yet their reliance on time-consuming multi-step sampling largely hinders their practical applications. While recent efforts have introduced few- or single-step solutions, existing methods either inefficiently model the process from noisy input or fail to fully exploit iterative generative priors, compromising the fidelity and quality of the reconstructed images. To address this issue, we propose FlowSR, a novel approach that reformulates the SR problem as a rectified flow from low-resolution (LR) to high-resolution (HR) images. Our method leverages an improved consistency learning strategy to enable high-quality SR in a single step. Specifically, we refine the original consistency distillation process by incorporating HR regularization, ensuring that the learned SR flow not only enforces self-consistency but also converges precisely to the ground-truth HR target. Furthermore, we introduce a fast-slow scheduling strategy, where adjacent timesteps for consistency learning are sampled from two distinct schedulers: a fast scheduler with fewer timesteps to improve efficiency, and a slow scheduler with more timesteps to capture fine-grained texture details. Extensive experiments demonstrate that FlowSR achieves outstanding performance in both efficiency and image quality. Code: \href{https://github.com/jiaqixuac/FlowSR}{this https URL}.

Figures

Figures reproduced from arXiv: 2605.12377 by Fan Li, Haoran Yang, Haoze Sun, Jiaqi Xu, Jingjing Ren, Long Peng, Pheng-Ann Heng, Renjing Pei, Wenbo Li, Xiaowei Hu, Zhixin Wang.

Figure 1
Figure 1. Figure 1: Our consistency SR flow model achieves high-quality [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Overview of the training process. The consistency SR flow distills multi-step, high-quality SR capability into single-step [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Illustration of HR-regularized consistency learning. Ap [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Visual comparisons of different SR methods on real-world examples. The number of sampling steps are indicated in parentheses. [PITH_FULL_IMAGE:figures/full_fig_p007_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Ablation of SR flow. Our SR flow model produces more [PITH_FULL_IMAGE:figures/full_fig_p007_5.png] view at source ↗
Figure 7
Figure 7. Figure 7: Ablation of training loss [PITH_FULL_IMAGE:figures/full_fig_p008_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Impact of timestep shifting / timestep sampling. SD3 [PITH_FULL_IMAGE:figures/full_fig_p011_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Visual comparisons of different SR methods on real-world examples #1. The number of sampling steps are indicated in bracket. [PITH_FULL_IMAGE:figures/full_fig_p013_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: Visual comparisons of different SR methods on real-world examples #2. The number of sampling steps are indicated in bracket. [PITH_FULL_IMAGE:figures/full_fig_p014_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: Visual comparisons of different SR methods on real-world examples #3. The number of sampling steps are indicated in bracket. [PITH_FULL_IMAGE:figures/full_fig_p015_11.png] view at source ↗

Review history (2 revisions) →

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

70 extracted references · 9 canonical work pages · 3 internal anchors

  1. [1]

    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 CVPRW, 2017. 11

    2017

  2. [2]

    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. InICCV, 2019. 6, 7

    2019

  3. [3]

    Adversarial diffu- sion compression for real-world image super-resolution

    Bin Chen, Gehui Li, Rongyuan Wu, Xindong Zhang, Jie Chen, Jian Zhang, and Lei Zhang. Adversarial diffu- sion compression for real-world image super-resolution. In CVPR, 2025. 3

    2025

  4. [4]

    Taming diffusion prior for image super-resolution with domain shift sdes.NeurIPS, 2024

    Qinpeng Cui, Yixuan Liu, Xinyi Zhang, Qiqi Bao, Zhong- dao Wang, Qingmin Liao, Li Wang, Tian Lu, and Emad Barsoum. Taming diffusion prior for image super-resolution with domain shift sdes.NeurIPS, 2024. 1, 3, 4, 6, 12

    2024

  5. [5]

    Image quality assessment: Unifying structure and texture similarity.TPAMI, 2020

    Keyan Ding, Kede Ma, Shiqi Wang, and Eero P Simoncelli. Image quality assessment: Unifying structure and texture similarity.TPAMI, 2020. 6

    2020

  6. [6]

    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. InECCV, 2014. 2

    2014

  7. [7]

    Tsd-sr: One-step diffusion with target score distillation for real-world image super-resolution

    Linwei Dong, Qingnan Fan, Yihong Guo, Zhonghao Wang, Qi Zhang, Jinwei Chen, Yawei Luo, and Changqing Zou. Tsd-sr: One-step diffusion with target score distillation for real-world image super-resolution. InCVPR, 2025. 3

    2025

  8. [8]

    Scaling recti- fied flow transformers for high-resolution image synthesis

    Patrick Esser, Sumith Kulal, Andreas Blattmann, Rahim Entezari, Jonas M ¨uller, Harry Saini, Yam Levi, Dominik Lorenz, Axel Sauer, Frederic Boesel, et al. Scaling recti- fied flow transformers for high-resolution image synthesis. InICML, 2024. 12

    2024

  9. [9]

    Generative adversarial nets.NeurIPS, 2014

    Ian Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron Courville, and Yoshua Bengio. Generative adversarial nets.NeurIPS, 2014. 2

    2014

  10. [10]

    One step diffusion-based super-resolution with time-aware distillation.arXiv preprint arXiv:2408.07476, 2024

    Xiao He, Huaao Tang, Zhijun Tu, Junchao Zhang, Kun Cheng, Hanting Chen, Yong Guo, Mingrui Zhu, Nan- nan Wang, Xinbo Gao, et al. One step diffusion-based super-resolution with time-aware distillation.arXiv preprint arXiv:2408.07476, 2024. 3

    work page arXiv 2024

  11. [11]

    Gans trained by a two time-scale update rule converge to a local nash equilib- rium.NeruIPS, 2017

    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.NeruIPS, 2017. 6

    2017

  12. [12]

    Denoising diffu- sion probabilistic models.NeruIPS, 2020

    Jonathan Ho, Ajay Jain, and Pieter Abbeel. Denoising diffu- sion probabilistic models.NeruIPS, 2020. 1, 2

    2020

  13. [13]

    Lora: Low-rank adaptation of large language models

    Edward J Hu, Yelong Shen, Phillip Wallis, Zeyuan Allen- Zhu, Yuanzhi Li, Shean Wang, Lu Wang, and Weizhu Chen. Lora: Low-rank adaptation of large language models. In ICLR, 2022. 6

    2022

  14. [14]

    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, 2019. 6

    2019

  15. [15]

    Musiq: Multi-scale image quality transformer

    Junjie Ke, Qifei Wang, Yilin Wang, Peyman Milanfar, and Feng Yang. Musiq: Multi-scale image quality transformer. InICCV, 2021. 6

    2021

  16. [16]

    Consistency trajectory mod- els: Learning probability flow ode trajectory of diffusion

    Dongjun Kim, Chieh-Hsin Lai, Wei-Hsiang Liao, Naoki Mu- rata, Yuhta Takida, Toshimitsu Uesaka, Yutong He, Yuki Mitsufuji, and Stefano Ermon. Consistency trajectory mod- els: Learning probability flow ode trajectory of diffusion. In ICLR, 2024. 3

    2024

  17. [17]

    Flux.https://github.com/ black-forest-labs/flux, 2024

    Black Forest Labs. Flux.https://github.com/ black-forest-labs/flux, 2024. 12

    2024

  18. [18]

    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. InCVPR, 2017. 2

    2017

  19. [19]

    Un- leashing the power of one-step diffusion based image super- resolution via a large-scale diffusion discriminator.arXiv preprint arXiv:2410.04224, 2024

    Jianze Li, Jiezhang Cao, Zichen Zou, Xiongfei Su, Xin Yuan, Yulun Zhang, Yong Guo, and Xiaokang Yang. Distillation-free one-step diffusion for real-world image super-resolution.arXiv preprint arXiv:2410.04224, 2024. 3

    work page arXiv 2024

  20. [20]

    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. InCVPR, 2023. 6

    2023

  21. [21]

    Swinir: Image restoration us- ing swin transformer

    Jingyun Liang, Jiezhang Cao, Guolei Sun, Kai Zhang, Luc Van Gool, and Radu Timofte. Swinir: Image restoration us- ing swin transformer. InICCV Workshop, 2021. 2

    2021

  22. [22]

    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. InECCV, 2024. 2, 6, 12

    2024

  23. [23]

    Flow matching for generative mod- eling

    Yaron Lipman, Ricky TQ Chen, Heli Ben-Hamu, Maximil- ian Nickel, and Matt Le. Flow matching for generative mod- eling. InICLR, 2023. 4

    2023

  24. [24]

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

    Xingchao Liu, Chengyue Gong, and Qiang Liu. Flow straight and fast: Learning to generate and transfer data with rectified flow. InICLR, 2023. 2, 3, 4

    2023

  25. [25]

    Instaflow: One step is enough for high-quality diffusion- based text-to-image generation

    Xingchao Liu, Xiwen Zhang, Jianzhu Ma, Jian Peng, et al. Instaflow: One step is enough for high-quality diffusion- based text-to-image generation. InICLR, 2024. 3

    2024

  26. [26]

    Knowledge Distillation in Iterative Generative Models for Improved Sampling Speed

    Eric Luhman and Troy Luhman. Knowledge distillation in iterative generative models for improved sampling speed. arXiv preprint arXiv:2101.02388, 2021. 2

    work page internal anchor Pith review Pith/arXiv arXiv 2021

  27. [27]

    Latent Consistency Models: Synthesizing High-Resolution Images with Few-Step Inference

    Simian Luo, Yiqin Tan, Longbo Huang, Jian Li, and Hang Zhao. Latent consistency models: Synthesizing high- resolution images with few-step inference.arXiv preprint arXiv:2310.04378, 2023. 3

    work page internal anchor Pith review Pith/arXiv arXiv 2023

  28. [28]

    completely blind

    Anish Mittal, Rajiv Soundararajan, and Alan C Bovik. Mak- ing a “completely blind” image quality analyzer.IEEE Sig- nal processing letters, 2012. 6

    2012

  29. [29]

    T2i-adapter: Learning adapters to dig out more controllable ability for text-to-image diffusion models

    Chong Mou, Xintao Wang, Liangbin Xie, Yanze Wu, Jian Zhang, Zhongang Qi, and Ying Shan. T2i-adapter: Learning adapters to dig out more controllable ability for text-to-image diffusion models. InAAAI, 2024. 2 9

    2024

  30. [30]

    You only need one step: Fast super-resolution with stable diffusion via scale distillation

    Mehdi Noroozi, Isma Hadji, Brais Martinez, Adrian Bulat, and Georgios Tzimiropoulos. You only need one step: Fast super-resolution with stable diffusion via scale distillation. InECCV, 2024. 3

    2024

  31. [31]

    Towards realis- tic data generation for real-world super-resolution

    Long Peng, Wenbo Li, Renjing Pei, Jingjing Ren, Jiaqi Xu, Yang Wang, Yang Cao, and Zheng-Jun Zha. Towards realis- tic data generation for real-world super-resolution. InICLR,

  32. [32]

    Learn- ing transferable visual models from natural language super- vision

    Alec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh, Gabriel Goh, Sandhini Agarwal, Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, et al. Learn- ing transferable visual models from natural language super- vision. InICML, 2021. 5

    2021

  33. [33]

    Ultrapixel: Advancing ultra high-resolution image synthesis to new peaks.NeurIPS, 2024

    Jingjing Ren, Wenbo Li, Haoyu Chen, Renjing Pei, Bin Shao, Yong Guo, Long Peng, Fenglong Song, and Lei Zhu. Ultrapixel: Advancing ultra high-resolution image synthesis to new peaks.NeurIPS, 2024. 2

    2024

  34. [34]

    High-resolution image syn- thesis with latent diffusion models

    Robin Rombach, Andreas Blattmann, Dominik Lorenz, Patrick Esser, and Bj¨orn Ommer. High-resolution image syn- thesis with latent diffusion models. InCVPR, 2022. 1, 2, 5, 6, 11

    2022

  35. [35]

    Image super- resolution via iterative refinement.TPAMI, 2022

    Chitwan Saharia, Jonathan Ho, William Chan, Tim Sali- mans, David J Fleet, and Mohammad Norouzi. Image super- resolution via iterative refinement.TPAMI, 2022. 2

    2022

  36. [36]

    Fast high- resolution image synthesis with latent adversarial diffusion distillation

    Axel Sauer, Frederic Boesel, Tim Dockhorn, Andreas Blattmann, Patrick Esser, and Robin Rombach. Fast high- resolution image synthesis with latent adversarial diffusion distillation. InSIGGRAPH Asia, 2024. 5, 12

    2024

  37. [37]

    Adversarial diffusion distillation

    Axel Sauer, Dominik Lorenz, Andreas Blattmann, and Robin Rombach. Adversarial diffusion distillation. InECCV, 2024. 2

    2024

  38. [38]

    Boosting latent diffusion with flow match- ing

    Johannes Schusterbauer, Ming Gui, Pingchuan Ma, Nick Stracke, Stefan Andreas Baumann, Vincent Tao Hu, and Bj¨orn Ommer. Boosting latent diffusion with flow match- ing. InECCV, 2024. 3, 4

    2024

  39. [39]

    Denois- ing diffusion implicit models

    Jiaming Song, Chenlin Meng, and Stefano Ermon. Denois- ing diffusion implicit models. InICLR, 2021. 1

    2021

  40. [40]

    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. InICLR, 2021. 1, 2

    2021

  41. [41]

    Consistency models

    Yang Song, Prafulla Dhariwal, Mark Chen, and Ilya Sutskever. Consistency models. InICML, 2023. 2, 3, 5, 8

    2023

  42. [42]

    Coser: Bridging image and language for cognitive super-resolution

    Haoze Sun, Wenbo Li, Jianzhuang Liu, Haoyu Chen, Ren- jing Pei, Xueyi Zou, Youliang Yan, and Yujiu Yang. Coser: Bridging image and language for cognitive super-resolution. InCVPR, 2024. 2

    2024

  43. [43]

    Phased consistency models.NeurIPS, 2024

    Fu-Yun Wang, Zhaoyang Huang, Alexander Bergman, Dazhong Shen, Peng Gao, Michael Lingelbach, Keqiang Sun, Weikang Bian, Guanglu Song, Yu Liu, et al. Phased consistency models.NeurIPS, 2024. 3, 5

    2024

  44. [44]

    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. In AAAI, 2023. 6

    2023

  45. [45]

    Exploiting diffusion prior for real-world image super-resolution.IJCV, 2024

    Jianyi Wang, Zongsheng Yue, Shangchen Zhou, Kelvin CK Chan, and Chen Change Loy. Exploiting diffusion prior for real-world image super-resolution.IJCV, 2024. 1, 2, 6, 11, 12

    2024

  46. [46]

    Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolution

    Peng Wang, Shuai Bai, Sinan Tan, Shijie Wang, Zhihao Fan, Jinze Bai, Keqin Chen, Xuejing Liu, Jialin Wang, Wenbin Ge, et al. Qwen2-vl: Enhancing vision-language model’s perception of the world at any resolution.arXiv preprint arXiv:2409.12191, 2024. 6

    work page internal anchor Pith review Pith/arXiv arXiv 2024

  47. [47]

    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. InICCV Workshop, 2021. 2, 6

    2021

  48. [48]

    Sinsr: diffusion-based image super- resolution in a single step

    Yufei Wang, Wenhan Yang, Xinyuan Chen, Yaohui Wang, Lanqing Guo, Lap-Pui Chau, Ziwei Liu, Yu Qiao, Alex C Kot, and Bihan Wen. Sinsr: diffusion-based image super- resolution in a single step. InCVPR, 2024. 1, 2, 6, 12

    2024

  49. [49]

    Prolificdreamer: High-fidelity and diverse text-to-3d generation with variational score distilla- tion.NeurIPS, 2023

    Zhengyi Wang, Cheng Lu, Yikai Wang, Fan Bao, Chongxuan Li, Hang Su, and Jun Zhu. Prolificdreamer: High-fidelity and diverse text-to-3d generation with variational score distilla- tion.NeurIPS, 2023. 2

    2023

  50. [50]

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

    Pengxu Wei, Ziwei Xie, Hannan Lu, Zongyuan Zhan, Qix- iang Ye, Wangmeng Zuo, and Liang Lin. Component divide-and-conquer for real-world image super-resolution. In ECCV, 2020. 6, 7, 11

    2020

  51. [51]

    One-step effective diffusion network for real-world image super-resolution.NeurIPS, 2024

    Rongyuan Wu, Lingchen Sun, Zhiyuan Ma, and Lei Zhang. One-step effective diffusion network for real-world image super-resolution.NeurIPS, 2024. 1, 2, 6, 12

    2024

  52. [52]

    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. InCVPR, 2024. 1, 2, 6, 12

    2024

  53. [53]

    Tack- ling the generative learning trilemma with denoising diffu- sion gans

    Zhisheng Xiao, Karsten Kreis, and Arash Vahdat. Tack- ling the generative learning trilemma with denoising diffu- sion gans. InICLR, 2022. 2

    2022

  54. [54]

    arXiv preprint arXiv:2404.01717 (2024)

    Rui Xie, Ying Tai, Chen Zhao, Kai Zhang, Zhenyu Zhang, Jun Zhou, Xiaoqian Ye, Qian Wang, and Jian Yang. Addsr: Accelerating diffusion-based blind super- resolution with adversarial diffusion distillation.arXiv preprint arXiv:2404.01717, 2024. 2

    work page arXiv 2024

  55. [55]

    Perflow: Piecewise rectified flow as universal plug-and-play accelerator.NeurIPS, 2024

    Hanshu Yan, Xingchao Liu, Jiachun Pan, Jun Hao Liew, Qiang Liu, and Jiashi Feng. Perflow: Piecewise rectified flow as universal plug-and-play accelerator.NeurIPS, 2024. 3

    2024

  56. [56]

    Consistency flow matching: Defining straight flows with velocity consistency.arXiv preprint arXiv:2407.02398, 2024

    Ling Yang, Zixiang Zhang, Zhilong Zhang, Xingchao Liu, Minkai Xu, Wentao Zhang, Chenlin Meng, Stefano Er- mon, and Bin Cui. Consistency flow matching: Defin- ing straight flows with velocity consistency.arXiv preprint arXiv:2407.02398, 2024. 3

    work page arXiv 2024

  57. [57]

    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. InCVPR, 2022. 6

    2022

  58. [58]

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

    Tao Yang, Rongyuan Wu, Peiran Ren, Xuansong Xie, and Lei Zhang. Pixel-aware stable diffusion for realistic im- age super-resolution and personalized stylization. InECCV,

  59. [59]

    Improved distribution matching distillation for fast image synthesis

    Tianwei Yin, Micha ¨el Gharbi, Taesung Park, Richard Zhang, Eli Shechtman, Fredo Durand, and Bill Freeman. Improved distribution matching distillation for fast image synthesis. NeurIPS, 2024. 5 10

    2024

  60. [60]

    One-step diffusion with distribution matching distillation

    Tianwei Yin, Micha ¨el Gharbi, Richard Zhang, Eli Shecht- man, Fredo Durand, William T Freeman, and Taesung Park. One-step diffusion with distribution matching distillation. In CVPR, 2024. 2

    2024

  61. [61]

    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. InCVPR, 2024. 1, 2

    2024

  62. [62]

    Resshift: Efficient diffusion model for image super- resolution by residual shifting.NeurIPS, 2023

    Zongsheng Yue, Jianyi Wang, and Chen Change Loy. Resshift: Efficient diffusion model for image super- resolution by residual shifting.NeurIPS, 2023. 1, 2, 4, 6, 7, 12

    2023

  63. [63]

    Arbitrary-steps image super-resolution via diffusion inver- sion

    Zongsheng Yue, Kang Liao, and Chen Change Loy. Arbitrary-steps image super-resolution via diffusion inver- sion. InCVPR, 2025. 3

    2025

  64. [64]

    Degradation-guided one-step im- age super-resolution with diffusion priors.arXiv preprint arXiv:2409.17058, 2024

    Aiping Zhang, Zongsheng Yue, Renjing Pei, Wenqi Ren, and Xiaochun Cao. Degradation-guided one-step im- age super-resolution with diffusion priors.arXiv preprint arXiv:2409.17058, 2024. 1, 3

    work page arXiv 2024

  65. [65]

    Designing a practical degradation model for deep blind image super-resolution

    Kai Zhang, Jingyun Liang, Luc Van Gool, and Radu Timo- fte. Designing a practical degradation model for deep blind image super-resolution. InICCV, 2021. 2

    2021

  66. [66]

    Adding conditional control to text-to-image diffusion models

    Lvmin Zhang, Anyi Rao, and Maneesh Agrawala. Adding conditional control to text-to-image diffusion models. In ICCV, 2023. 2

    2023

  67. [67]

    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. InCVPR, 2018. 5, 6 In this supplementary material, we first provide addi- tional details about our FlowSR in Sec. 7. Next, we present more experimental results in Sec. 8. Finally, we discuss the limitations ...

    2018

  68. [68]

    The fine-tuned SR flow model is then used to initialize both the SR modelθand the teacher modelϕ

    Implementation Details We first fine-tune the pre-trained SD model [34] to adapt it to our SR flow learning objectives. The fine-tuned SR flow model is then used to initialize both the SR modelθand the teacher modelϕ. A default text prompt is used for the SD model. During consistency SR flow training, each train- ing batch is split into two groups: one fo...

  69. [69]

    Evaluation on DIV2K-Val We also evaluate our method on the DIV2K-Val dataset [1, 45]

    More Results 8.1. Evaluation on DIV2K-Val We also evaluate our method on the DIV2K-Val dataset [1, 45]. Table 6 provides a quantitative comparison of var- ious SR methods. Across all reference-based metrics, our FlowSR achieves state-of-the-art performance or performs on par with the best existing methods. For no-reference metrics, while FlowSR performs w...

    work page arXiv 1968

  70. [70]

    We provide valuable insights into the effective use of flow-based techniques and consistency learning to achieve competitive SR results in a single-step setting

    Limitations and Future Works In this work, we tackle one-step SR from the perspective of flow and consistency. We provide valuable insights into the effective use of flow-based techniques and consistency learning to achieve competitive SR results in a single-step setting. While our study demonstrates promising results, there are some limitations. First, d...