Pith. sign in

REVIEW 4 major objections 7 minor 67 references

This paper argues that single-number correlation metrics mislead IQA benchmarking, and introduces a correlation surface and a distribution-robust global score.

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 ↗

GMC maps an IQA model's agreement with human scores across the quality-level and quality-difference landscape, exposing local strengths that global PLCC/SRCC hide.

T0 review reviewed 2026-08-03 challenge →

load-bearing objection GMC is a genuinely useful diagnostic surface for IQA evaluation — worth engaging seriously — but the 'distribution-agnostic' claim is not yet supported; it needs a bias analysis before GMCg can be treated as a standard metric. the 4 major comments →

arxiv 2601.21738 v2 pith:HJJOQQKY submitted 2026-01-29 cs.CV cs.AI

From Global to Granular: Revealing IQA Model Performance via Correlation Surface

Baoliang Chen , Danni Huang , Hanwei Zhu , Lingyu Zhu , Wei Zhou , Shiqi Wang , Yuming Fang , Weisi Lin This is my paper
classification cs.CV cs.AI
keywords image quality assessmentcorrelation coefficientfine-grained evaluationdistribution robustnessmodel benchmarkingrank correlationperceptual metrics
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 reading

Image quality assessment (IQA) models are usually compared by a single number — PLCC, SRCC, or KRCC — computed over a whole benchmark. The paper argues that this scalar hides which quality regimes a model actually handles: two models can tie on SRCC while one is accurate on high-quality images and the other is better at telling apart nearly identical distortions. It proposes Granularity-Modulated Correlation (GMC), a weighted correlation that can be queried at any combination of absolute quality (MOS) and pairwise quality difference (|ΔMOS|), then assembled into a 3D correlation surface. Averaging that surface gives a global score, GMC_g, which the paper claims is stable when the test set's quality distribution changes, unlike SRCC or PLCC. If true, GMC would give researchers a way to rank models fairly and to choose models for specific applications, such as evaluating generative outputs versus guiding compression.

Core claim

The central claim is that IQA performance is a two-dimensional function — prediction accuracy along absolute MOS and discrimination capability along pairwise MOS difference — and that both dimensions can be measured with a generalized correlation coefficient whose per-pair weights come from two modulators: a Granularity Modulator that down-weights pairs far from a queried (MOS, |ΔMOS|) via Gaussians with per-image rating variance, and a Distribution Regulator that reweights pairs by the inverse of a kernel-smoothed MOS density. Sampling these weighted correlations at Latin-Hypercube points and fitting a local-linear kernel regression yields a continuous correlation surface; integrating the s

What carries the argument

The machinery is the Generalized Correlation Coefficient (GCC) extended with a per-pair weight w^k_ij = P^s_k(i,j) * P^d_k(i,j) * P^t_k(i,j). P^s and P^d form the Granularity Modulator: Gaussian probabilities that images i and j have MOS about Q^s_k and pairwise difference about Q^d_k, using each image's rating standard deviation σ_i. P^t is the Distribution Regulator: the inverse of kernel-smoothed density at each image's MOS. Replacing GCC's uniform pairing with these weights turns PLCC, SRCC, and KRCC into local probes; a Latin Hypercube design samples the (Q^s, Q^d) plane, local linear kernel regression fits the surface, and surface integration produces the global GMC_g.

Load-bearing premise

The entire surface and the stability of GMC_g rest on treating each image's MOS as a Gaussian random variable with known variance σ_i^2, so that the granularity weights in Eqs. (13)–(15) faithfully represent perceptual relevance; if σ_i is missing or the Gaussian model is wrong, every local score inherits that error.

What would settle it

Measure GMC on a dataset whose rating noise is deliberately non-Gaussian (e.g., heavy-tailed or bimodal within-image ratings) while keeping the same MOS values; if the surface and GMC_g change materially, the Gaussian weighting is the load-bearing assumption. A simpler calculation: replace σ_i in Eqs. (13)–(15) with a constant global σ across images and check whether conclusions about regime-specific strengths (e.g., CLIP-IQA vs NIQE) invert—if they do, the per-image σ estimates are driving the result, not the data.

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

If this is right

  • Two IQA models with identical global SRCC can be separated by their correlation surfaces, enabling benchmarks to report regime-specific strengths instead of one ambiguous number.
  • Model selection becomes scenario-aware: the surface identifies models suited to high-MOS/fine-difference tasks (e.g., generative image evaluation) versus low-MOS/coarse-difference tasks (e.g., heavy compression).
  • GMC_g provides a global ranking that is more stable than SRCC when the test set's quality distribution shifts, making cross-dataset and re-sampled comparisons fairer.
  • The finding that all tested models degrade sharply at small |ΔMOS| indicates a concrete open problem: current IQA models lack fine-grained discrimination, and GMC makes that gap visible and measurable.
  • GMC-guided integration of complementary models (e.g., MS-SSIM with LPIPS) outperforms integration chosen by SRCC, suggesting the surface can drive ensembles.

Where Pith is reading between the lines

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

  • The same weighted-correlation surface could be ported to other perceptual or non-perceptual evaluation settings where a scalar metric hides local behavior, such as video quality, text-to-image alignment, or ranking models in information retrieval.
  • GMC_g's distribution-agnostic claim depends on the Gaussian weighting surviving non-Gaussian rating noise; if σ_i is badly estimated (e.g., via Beta regression on sparse ratings), a non-parametric or heteroscedastic weighting would be a natural robustness check.
  • The surface could be inverted into a training signal: penalizing a model's low-difference-regime GMC during training might directly improve fine-grained discrimination rather than global SRCC.
  • Because GMC_g integrates over the whole MOS/ΔMOS domain, it effectively requires a test set spanning the full quality range; applying it to truncated or narrow-range datasets would need explicit domain extension, which the paper does not address.
Share X Bluesky LinkedIn Reddit HN

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

4 major / 7 minor

Summary. The paper proposes Granularity-Modulated Correlation (GMC), a generalization of Kendall's Generalized Correlation Coefficient (GCC) in which pair weights depend on absolute MOS and pairwise MOS difference through Gaussian kernels (Granularity Modulator), and are inversely weighted by a kernel-smoothed quality density (Distribution Regulator). Local weighted correlations are sampled via Latin hypercube sampling and fit by local linear regression to produce a 3D correlation surface over (MOS, |ΔMOS|); integrating this surface yields a global score GMCg. The paper reports experiments on KADID-10k, PIPAL, LIVEC, and SPAQ with both FR and NR IQA models, and uses GMC for local diagnostics, scenario-specific model selection, fine-grained optimization, model integration, and robustness to sampling distributions. The mathematical extension of GCC with nonnegative symmetric weights is sound, and the local surface construction is clearly described, but the strongest claim—that GMCg is an "inherently distribution-agnostic" and "true" reflection of model capability—is not supported by the current evidence.

Significance. If validated, GMC would be a useful diagnostic tool for the IQA community: it localizes correlation performance in a way global PLCC/SRCC cannot, and the surface visualization plus GMCg could support model selection and integration. The authors provide clean derivations linking PLCC, SRCC, and KRCC to GCC, and they release code, which aids reproducibility. The core novelty is moderate but potentially valuable, especially for applications requiring fine-grained discrimination. However, the most ambitious claim—distribution-agnostic and unbiased global scoring—needs substantially more evidence, and some validation experiments are qualitative or statistically thin. With additional bias analysis, confidence intervals, and a corrected interpretation of the Gaussian weights, the contribution could be solid.

major comments (4)
  1. [§3.3.1, Eqs. (13)–(15)] The Gaussian weighting is presented as a probability model, but Eqs. (13)–(15) conflate the MOS estimate q_i with the random subjective rating R_i. If R_i ~ N(q_i, σ_i^2), then σ_i is the spread of individual ratings, not the uncertainty of the mean q_i; Eq. (13) is not the probability that q_i equals Q_s. Moreover, if R_i and R_j are independent Gaussians, |R_i − R_j| follows a folded normal distribution, not N(|q_i − q_j|, σ_i^2 + σ_j^2), so Eq. (15) is not a valid probability for |q_i − q_j| = Q_d. Since these weights define every localized correlation and all GMC_s/GMC_d conclusions, the manuscript must either provide a correct probabilistic derivation or explicitly re-frame Eqs. (13)–(15) as heuristic kernel weighting and analyze sensitivity to the choice of σ_i (and to Beta-regression estimates when σ_i is unavailable).
  2. [§3.3.3, Eq. (22); §4.2(5)] The claim that GMCg is "inherently distribution-agnostic" and yields a "true" reflection of model capability is not established. GMCg integrates the fitted surface over the observed rectangle [Q_s_min, Q_s_max] × [Q_d_min, Q_d_max]. The surface is estimated from a single sample: inverse-density weighting in Eqs. (16)–(18) cannot create information in unsupported MOS/|ΔMOS| regions, where local linear regression must interpolate or extrapolate. Section 4.2(5) reports only the standard deviation of GMCg across nine Gaussian-shaped subsets; low variance can coexist with large systematic bias. No comparison to the full-dataset GMCg, no bias/coverage analysis, and no distribution-robust baseline (e.g., PWRC) are provided. The authors should add a bias analysis and confidence intervals, and temper the "true reflection" claim accordingly.
  3. [Table 3, §4.3] The ablation of kernel-smoothed density estimation does not support the stated conclusion. Across five models on two datasets, the reduction in standard deviation from raw bin-frequency density is on the order of 1e-4 to 1e-5, and on SPAQ CLIP-IQA the kernel variant has higher std (0.007805 vs 0.007800). No significance test or effect on model ranking is reported. Since the Distribution Regulator is a core module, this experiment needs to be redesigned—for example, with more distribution shifts, repeated sampling, and reporting bias as well as variance—or the conclusion must be weakened.
  4. [§4.2(3), Fig. 7] The adversarial optimization experiment is presented as confirmation that MS-SSIM is superior in the low-difference regime, but the evidence is qualitative: four example images per setting, no quantitative metrics for constraint satisfaction (e.g., the fraction of runs where the adversarial metric remains within ε), no average improvements in the target metric, and no human evaluation of the optimized images. Please provide quantitative summaries and statistical comparisons; otherwise the link between the GMC surface and optimization utility remains anecdotal.
minor comments (7)
  1. [Eqs. (19)–(20)] Both sampling coordinates use the same u_k ~ U(0,1). Standard LHS requires independent perturbations u_k^x and u_k^y; using the same draw induces a correlation between Q_s and Q_d sample coordinates.
  2. [§4.1] The bandwidths for the kernel density estimator in Eqs. (17)–(18) and for the local linear regression in Eq. (21) are not specified, and the Beta-regression procedure for estimating σ_i is not described. These are important free parameters for reproducibility.
  3. [Table 3] The model names QALIGN and TOPIQ-NR appear in Table 3 but are not among the NR models introduced in Section 4.1; please define them and align the table with the model list.
  4. [Table 1] DISTS is cited as [31] in Table 1, but the correct reference is [5]. Check also whether references [60] and [22] are duplicated.
  5. [Fig. 7] The caption contains a typo: "PIAPL" should be "PIPAL".
  6. [Eq. (18)] The density estimator is written as an integral over Y, but Y is elsewhere described as a set of bins; please clarify the notation and the kernel bandwidth.
  7. [Figs. 10–11] The axes are not labeled, making it hard to verify the convergence claim. Add axis labels and a legend.

Circularity Check

0 steps flagged

No significant circularity: GMC is an a priori weighting scheme and its validation uses independent objectives.

full rationale

The derivation chain is self-contained. GMC is obtained by inserting a fixed, a priori weighting scheme—Gaussian weights in MOS and |ΔMOS| plus inverse-density regularization—into the GCC formula (Eqs. 8–18). No parameter is fitted to the model rankings or to any target outcome, so there is no fitted input masquerading as a prediction. The local diagnostics GMC_s/GMC_d and the global GMC_g are definitions (Eqs. 22–24), not predictions, so reading performance off the constructed surface is not circular. The validation experiments use objectives independent of GMC: model selection is checked against the actual mean MOS of retrieved images, integration is checked with the SRCC of the combined metric, and adversarial optimization uses MS-SSIM/LPIPS as the objectives with GMC only generating the hypothesis. The robustness experiment in Sec. 4.2(5) demonstrates low variance across shifted subsets but does not establish unbiasedness; that is an evidentiary gap in the 'true reflection' claim, not a circular reduction. Self-citations exist (e.g., [24], [33], [34], [47], [48]), but they are not load-bearing: the Gaussian rating assumption is also supported by external references [55], [56], and no uniqueness theorem or forced choice rests on the authors' own prior work. Accordingly, no specific circular step can be exhibited.

Axiom & Free-Parameter Ledger

4 free parameters · 3 axioms · 0 invented entities

GMC introduces no new physical or ontological entities; it is a statistical evaluation measure. Its load-bearing premises are the Gaussianity and independence assumptions for human ratings, which are domain assumptions without direct empirical validation in the paper. Free parameters include the per-image std (when not supplied by the dataset) and the kernel bandwidths for density estimation and surface regression, none of which are reported or analyzed for sensitivity.

free parameters (4)
  • Per-image rating standard deviation σ_i = Dataset-dependent; not reported; estimated via Beta regression when missing
    Used in Eqs. (13) and (15) to define Gaussian weights for MOS and |ΔMOS|; controls the width of the local 'attention' and is load-bearing for the granularity analysis.
  • Kernel bandwidth for density estimator D(q) = Not specified
    Eqs. (17)-(18) use Gaussian kernels without specifying bandwidth; the distribution regulator's upweighting depends on this choice.
  • Kernel bandwidth for local linear regression surface fit = Not specified
    Eq. (21) fits the surface with local linear kernel regression; the bandwidth controls smoothness and affects GMCg via integration in Eq. (22).
  • Number of LHS sample points K and bin count Y = K=100, Y=100
    Set in Sec. 4.1; ablation varies K but not Y; these choices affect surface sampling and integration ranges.
axioms (3)
  • domain assumption Human quality ratings for each image are Gaussian-distributed with mean MOS qi and standard deviation σ_i
    Invoked in Sec. 3.3.1 to justify the Gaussian forms of Ps(i,j) and Pd(i,j) weights (Eqs. 13-15).
  • domain assumption MOS values of different images are statistically independent, allowing the joint probability to factor into a product of marginals
    Stated in Eq. (14); needed for the product form of the granularity weights.
  • domain assumption The pair-wise quality difference |qi - qj| follows a Gaussian with variance σ_i^2 + σ_j^2
    Used implicitly in Eq. (15); assumes independence of rating errors for the two images.

reviewed 2026-08-03 · how reviews work

0 comments
Cite this review

Pith. "Pith review of From Global to Granular: Revealing IQA Model Performance via Correlation Surface." pith.science (2026). https://pith.science/paper/HJJOQQKY

@misc{pith2026260121738,
  author       = {Pith},
  title        = {Pith review of: From Global to Granular: Revealing IQA Model Performance via Correlation Surface},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/HJJOQQKY}},
  note         = {Machine review of arXiv:2601.21738}
}
Share X Bluesky LinkedIn Reddit HN
abstract

Evaluation of Image Quality Assessment (IQA) models has long been dominated by global correlation metrics, such as Pearson Linear Correlation Coefficient (PLCC) and Spearman Rank-Order Correlation Coefficient (SRCC). While widely adopted, these metrics reduce performance to a single scalar, failing to capture how ranking consistency varies across the local quality spectrum. For example, two IQA models may achieve identical SRCC values, yet one ranks high-quality images (related to high Mean Opinion Score, MOS) more reliably, while the other better discriminates image pairs with small quality/MOS differences (related to $|\Delta$MOS$|$). Such complementary behaviors are invisible under global metrics. Moreover, SRCC and PLCC are sensitive to test-sample quality distributions, yielding unstable comparisons across test sets. To address these limitations, we propose \textbf{Granularity-Modulated Correlation (GMC)}, which provides a structured, fine-grained analysis of IQA performance. GMC includes: (1) a \textbf{Granularity Modulator} that applies Gaussian-weighted correlations conditioned on absolute MOS values and pairwise MOS differences ($|\Delta$MOS$|$) to examine local performance variations, and (2) a \textbf{Distribution Regulator} that regularizes correlations to mitigate biases from non-uniform quality distributions. The resulting \textbf{correlation surface} maps correlation values as a joint function of MOS and $|\Delta$MOS$|$, providing a 3D representation of IQA performance. Experiments on standard benchmarks show that GMC reveals performance characteristics invisible to scalar metrics, offering a more informative and reliable paradigm for analyzing, comparing, and deploying IQA models. Codes are available at https://github.com/Dniaaa/GMC.

Figures

Figures reproduced from arXiv: 2601.21738 by Baoliang Chen, Danni Huang, Hanwei Zhu, Lingyu Zhu, Shiqi Wang, Weisi Lin, Wei Zhou, Yuming Fang.

Figure 1
Figure 1. Figure 1: Complementary IQA behaviors along two coupled assessment dimensions on the SPAQ dataset [20]. The overall performance of an IQA model is governed by its prediction accuracy relative to absolute quality (MOS) and its discrimination capability regarding pairwise differences (|∆MOS|). (a) Performance snapshots: Images from the SPAQ dataset are partitioned into four subsets based on the medians of MOS and |∆MO… view at source ↗
Figure 2
Figure 2. Figure 2: Sensitivity of global correlation metrics (PLCC and SRCC) to [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Overview of the proposed 3D GMC performance surface for fine-grained IQA evaluation. The framework consists of a [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Density estimation at a target quality score [PITH_FULL_IMAGE:figures/full_fig_p006_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Visualization of correlation surfaces generated by different IQA models. The “GMC” response (vertical axis) is shown as a function of the [PITH_FULL_IMAGE:figures/full_fig_p008_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Comparison of high-quality images retrieved by the best FR/NR IQA models selected via GMC and the SRCC criterion. (a) and (b) show the [PITH_FULL_IMAGE:figures/full_fig_p010_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Qualitative comparison of adversarial image quality optimization on the PIAPL Dataset. Columns 1 and 2 show the reference and initial [PITH_FULL_IMAGE:figures/full_fig_p011_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Comparison of IQA model integration results. (a) and (c) show [PITH_FULL_IMAGE:figures/full_fig_p011_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Robustness comparison between SRCC and GMC [PITH_FULL_IMAGE:figures/full_fig_p012_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: Variation of GMCg across IQA models versus sampling size under Latin Hypercube Sampling [PITH_FULL_IMAGE:figures/full_fig_p013_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: Variation of GMCg across IQA models versus sampling size under random sampling. TABLE 3 Effect of kernel smoothing on GMCg variance across IQA models. Kernel Model Standard Devia￾tion Average PIPAL ✗ SSIM 0.010886 0.004167 PSNR 0.003117 MS-SSIM 0.002358 LPIPS 0.002999 DISTS 0.001477 ✓ SSIM 0.010804 0.004119 PSNR 0.003041 MS-SSIM 0.002316 LPIPS 0.002929 DISTS 0.001505 SPAQ ✗ CLIPIQA 0.007800 0.003227 CLIPI… view at source ↗

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

67 extracted references · 5 linked inside Pith

  1. [1]

    Modern image quality assessment,

    Z. Wang and A. C. Bovik, “Modern image quality assessment,” in Synthesis Lectures on Image, Video, and Multimedia Processing, 2006, pp. 1–156

    2006

  2. [2]

    Perceptual image quality assessment: a survey,

    G. Zhai and X. Min, “Perceptual image quality assessment: a survey,”Science China Information Sciences, vol. 63, pp. 1–52, 2020

    2020

  3. [3]

    Comparison of full-reference image quality models for optimization of image processing systems,

    K. Ding, K. Ma, S. Wang, and E. P . Simoncelli, “Comparison of full-reference image quality models for optimization of image processing systems,”International Journal of Computer Vision, vol. 129, no. 4, pp. 1258–1281, 2021

    2021

  4. [4]

    The effects of a visual fidelity criterion of the encoding of images,

    J. Mannos and D. Sakrison, “The effects of a visual fidelity criterion of the encoding of images,”IEEE Transactions on Information Theory, vol. 20, no. 4, pp. 525–536, 1974. SUBMITTED TO IEEE TRANSACTIONS ON PATTERN ANAL YSIS AND MACHINE INTELLIGENCE 14

    1974

  5. [5]

    Image quality assessment: Unifying structure and texture similarity,

    K. Ding, K. Ma, S. Wang, and E. P . Simoncelli, “Image quality assessment: Unifying structure and texture similarity,”IEEE Trans- actions on Pattern Analysis and Machine Intelligence, vol. 44, no. 5, pp. 2567–2581, 2020

    2020

  6. [6]

    Mean squared error: Love it or leave it? A new look at signal fidelity measures,

    Z. Wang and A. C. Bovik, “Mean squared error: Love it or leave it? A new look at signal fidelity measures,”IEEE Signal Processing Magazine, vol. 26, no. 1, pp. 98–117, 2009

    2009

  7. [7]

    What’s wrong with mean-squared error?

    B. Girod, “What’s wrong with mean-squared error?” inDigital images and human vision, 1993, pp. 207–220

    1993

  8. [8]

    Blind image quality assessment using a deep bilinear convolutional neural network,

    W. Zhang, K. Ma, J. Yan, D. Deng, and Z. Wang, “Blind image quality assessment using a deep bilinear convolutional neural network,”IEEE Transactions on Circuits and Systems for Video Tech- nology, vol. 30, no. 1, pp. 36–47, 2018

    2018

  9. [9]

    Deep CNN-based blind im- age quality predictor,

    J. Kim, A.-D. Nguyen, and S. Lee, “Deep CNN-based blind im- age quality predictor,”IEEE Transactions on Neural Networks and Learning Systems, vol. 30, no. 1, pp. 11–24, 2018

    2018

  10. [10]

    Deep neural networks for no-reference and full-reference image quality assessment,

    S. Bosse, D. Maniry, K.-R. M ¨uller, T. Wiegand, and W. Samek, “Deep neural networks for no-reference and full-reference image quality assessment,”IEEE Transactions on Image Processing, vol. 27, no. 1, pp. 206–219, 2017

    2017

  11. [11]

    TOPIQ: A top-down approach from semantics to distortions for image quality assessment,

    C. Chen, J. Mo, J. Hou, H. Wu, L. Liao, X. Min, and W. Lin, “TOPIQ: A top-down approach from semantics to distortions for image quality assessment,”IEEE Transactions on Image Processing, 2024

    2024

  12. [12]

    Transformer for image quality assess- ment,

    J. You and J. Korhonen, “Transformer for image quality assess- ment,” inIEEE International Conference on Image Processing, 2021, pp. 1389–1393

    2021

  13. [13]

    Perceptual image quality assessment with transformers,

    M. Cheon, S.-J. Yoon, B. Kang, and J. Lee, “Perceptual image quality assessment with transformers,” inIEEE/CVF Conference on Computer Vision and Pattern Recognition, 2021, pp. 433–442

    2021

  14. [14]

    MUSIQ: Multi-scale image quality transformer,

    J. Ke, Q. Wang, Y. Wang, P . Milanfar, and F. Yang, “MUSIQ: Multi-scale image quality transformer,” inIEEE/CVF International Conference on Computer Vision, 2021, pp. 5148–5157

    2021

  15. [15]

    Blind image qual- ity assessment via vision-language correspondence: A multitask learning perspective,

    W. Zhang, G. Zhai, Y. Wei, X. Yang, and K. Ma, “Blind image qual- ity assessment via vision-language correspondence: A multitask learning perspective,” inIEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 14 071–14 081

    2023

  16. [16]

    Boosting image quality assessment through efficient transformer adaptation with local feature enhancement,

    K. Xu, L. Liao, J. Xiao, C. Chen, H. Wu, Q. Yan, and W. Lin, “Boosting image quality assessment through efficient transformer adaptation with local feature enhancement,” inIEEE/CVF Confer- ence on Computer Vision and Pattern Recognition, 2024, pp. 2662– 2672

    2024

  17. [17]

    Q-Bench: A benchmark for general- purpose foundation models on low-level vision,

    H. Wu, Z. Zhang, E. Zhang, C. Chen, L. Liao, A. Wang, C. Li, W. Sun, Q. Yan, G. Zhaiet al., “Q-Bench: A benchmark for general- purpose foundation models on low-level vision,” inInternational Conference on Learning Representations, 2024

    2024

  18. [18]

    VisualQuality-R1: Reasoning-induced image quality assessment via reinforcement learning to rank,

    T. Wu, J. Zou, J. Liang, L. Zhang, and K. Ma, “VisualQuality-R1: Reasoning-induced image quality assessment via reinforcement learning to rank,”arXiv preprint arXiv:2505.14460, 2025

    arXiv 2025

  19. [19]

    Fine-grained image quality assessment: A revisit and further thinking,

    X. Zhang, W. Lin, and Q. Huang, “Fine-grained image quality assessment: A revisit and further thinking,”IEEE Transactions on Circuits and Systems for Video Technology, vol. 32, no. 5, pp. 2746– 2759, 2021

    2021

  20. [20]

    Perceptual quality assessment of smartphone photography,

    Y. Fang, H. Zhu, Y. Zeng, K. Ma, and Z. Wang, “Perceptual quality assessment of smartphone photography,” inIEEE/CVF Conference on Computer Vision and Pattern Recognition, 2020, pp. 3677–3686

    2020

  21. [21]

    Exploring CLIP for assessing the look and feel of images,

    J. Wang, K. C. Chan, and C. C. Loy, “Exploring CLIP for assessing the look and feel of images,” inAAAI Conference on Artificial Intelligence, vol. 37, no. 2, 2023, pp. 2555–2563

    2023

  22. [22]

    Making a “com- pletely blind

    A. Mittal, R. Soundararajan, and A. C. Bovik, “Making a “com- pletely blind” image quality analyzer,”IEEE Signal Processing Letters, vol. 20, no. 3, pp. 209–212, 2012

    2012

  23. [23]

    J. rank correlation methods,

    M. Kendall and D. Gibbons, “J. rank correlation methods,” 1990

    1990

  24. [24]

    Blind image quality assessment by gaussian mixture distribution,

    Y. Gao, X. Min, Y. Cao, W. Lin, B. S. Lee, and G. Zhai, “Blind image quality assessment by gaussian mixture distribution,”IEEE Transactions on Image Processing, 2025

    2025

  25. [25]

    Image quality assessment: from error visibility to structural similarity,

    Z. Wang, A. C. Bovik, H. R. Sheikh, and E. P . Simoncelli, “Image quality assessment: from error visibility to structural similarity,” IEEE transactions on image processing, vol. 13, no. 4, pp. 600–612, 2004

    2004

  26. [26]

    Multiscale structural similarity for image quality assessment,

    Z. Wang, E. P . Simoncelli, and A. C. Bovik, “Multiscale structural similarity for image quality assessment,” inAsilomar Conference on Signals, Systems & Computers, 2003, pp. 1398–1402

    2003

  27. [27]

    Information content weighting for perceptual image quality assessment,

    Z. Wang and Q. Li, “Information content weighting for perceptual image quality assessment,”IEEE Transactions on Image Processing, vol. 20, no. 5, pp. 1185–1198, 2010

    2010

  28. [28]

    Image information and visual quality,

    H. R. Sheikh and A. C. Bovik, “Image information and visual quality,”IEEE Transactions on image processing, vol. 15, no. 2, pp. 430–444, 2006

    2006

  29. [29]

    FSIM: A feature similarity index for image quality assessment,

    L. Zhang, L. Zhang, X. Mou, and D. Zhang, “FSIM: A feature similarity index for image quality assessment,”IEEE transactions on Image Processing, vol. 20, no. 8, pp. 2378–2386, 2011

    2011

  30. [30]

    Gradient magnitude similarity deviation: A highly efficient perceptual image quality index,

    W. Xue, L. Zhang, X. Mou, and A. C. Bovik, “Gradient magnitude similarity deviation: A highly efficient perceptual image quality index,”IEEE Transactions on Image Processing, vol. 23, no. 2, pp. 684–695, 2013

    2013

  31. [31]

    The unreasonable effectiveness of deep features as a perceptual met- ric,

    R. Zhang, P . Isola, A. A. Efros, E. Shechtman, and O. Wang, “The unreasonable effectiveness of deep features as a perceptual met- ric,” inIEEE Conference on Computer Vision and Pattern Recognition, 2018, pp. 586–595

    2018

  32. [32]

    Deep- WSD: Projecting degradations in perceptual space to wasserstein distance in deep feature space,

    X. Liao, B. Chen, H. Zhu, S. Wang, M. Zhou, and S. Kwong, “Deep- WSD: Projecting degradations in perceptual space to wasserstein distance in deep feature space,” inACM International Conference on Multimedia, 2022, pp. 970–978

    2022

  33. [33]

    DeepDC: Deep distance correlation as a perceptual image quality evaluator,

    H. Zhu, B. Chen, L. Zhu, S. Wang, and W. Lin, “DeepDC: Deep distance correlation as a perceptual image quality evaluator,” arXiv preprint arXiv:2211.04927, 2022

    Pith/arXiv arXiv 2022

  34. [34]

    Debi- ased mapping for full-reference image quality assessment,

    B. Chen, H. Zhu, L. Zhu, S. Wang, J. Pan, and S. Wang, “Debi- ased mapping for full-reference image quality assessment,”IEEE Transactions on Multimedia, 2025

    2025

  35. [35]

    A two-step framework for constructing blind image quality indices,

    A. K. Moorthy and A. C. Bovik, “A two-step framework for constructing blind image quality indices,”IEEE Signal Processing Letters, vol. 17, no. 5, pp. 513–516, 2010

    2010

  36. [36]

    Blind image quality assessment: A natural scene statistics approach in the dct domain,

    M. A. Saad, A. C. Bovik, and C. Charrier, “Blind image quality assessment: A natural scene statistics approach in the dct domain,” IEEE Transactions on Image Processing, vol. 21, no. 8, pp. 3339–3352, 2012

    2012

  37. [37]

    A psychovisual quality metric in free-energy principle,

    G. Zhai, X. Wu, X. Yang, W. Lin, and W. Zhang, “A psychovisual quality metric in free-energy principle,”IEEE Transactions on Image Processing, vol. 21, no. 1, pp. 41–52, 2011

    2011

  38. [38]

    No-reference image quality assessment metric by combining free energy theory and structural degradation model,

    K. Gu, G. Zhai, X. Yang, W. Zhang, and L. Liang, “No-reference image quality assessment metric by combining free energy theory and structural degradation model,” inIEEE International Conference on Multimedia and Expo, 2013, pp. 1–6

    2013

  39. [39]

    No- reference image quality assessment by hallucinating pristine fea- tures,

    B. Chen, L. Zhu, C. Kong, H. Zhu, S. Wang, and Z. Li, “No- reference image quality assessment by hallucinating pristine fea- tures,”IEEE Transactions on Image Processing, vol. 31, pp. 6139– 6151, 2022

    2022

  40. [40]

    Convolutional neural networks for no-reference image quality assessment,

    L. Kang, P . Ye, Y. Li, and D. Doermann, “Convolutional neural networks for no-reference image quality assessment,” inIEEE/CVF Conference on Computer Vision and Pattern Recognition, 2014, pp. 1733–1740

    2014

  41. [41]

    Blindly assess image quality in the wild guided by a self-adaptive hyper network,

    S. Su, Q. Yan, Y. Zhu, C. Zhang, X. Ge, J. Sun, and Y. Zhang, “Blindly assess image quality in the wild guided by a self-adaptive hyper network,” inIEEE/CVF Conference on Computer Vision and Pattern Recognition, June 2020

    2020

  42. [42]

    MetaIQA: Deep meta- learning for no-reference image quality assessment,

    H. Zhu, L. Li, J. Wu, W. Dong, and G. Shi, “MetaIQA: Deep meta- learning for no-reference image quality assessment,” inIEEE/CVF Conference on Computer Vision and Pattern Recognition, 2020, pp. 14 143–14 152

    2020

  43. [43]

    Task-specific normaliza- tion for continual learning of blind image quality models,

    W. Zhang, K. Ma, G. Zhai, and X. Yang, “Task-specific normaliza- tion for continual learning of blind image quality models,”IEEE Transactions on Image Processing, 2024

    2024

  44. [44]

    Learning generalized spatial-temporal deep feature representation for no- reference video quality assessment,

    B. Chen, L. Zhu, G. Li, F. Lu, H. Fan, and S. Wang, “Learning generalized spatial-temporal deep feature representation for no- reference video quality assessment,”IEEE Transactions on Circuits and Systems for Video Technology, vol. 32, no. 4, pp. 1903–1916, 2021

    1903

  45. [45]

    Q-Align: Teaching LMMs for visual scoring via discrete text-defined levels,

    H. Wu, Z. Zhang, W. Zhang, C. Chen, L. Liao, C. Li, Y. Gao, A. Wang, E. Zhang, W. Sunet al., “Q-Align: Teaching LMMs for visual scoring via discrete text-defined levels,”arXiv preprint arXiv:2312.17090, 2023

    Pith/arXiv arXiv 2023

  46. [46]

    Dog-IQA: Standard-guided zero-shot mllm for mix-grained image quality assessment,

    K. Liu, Z. Zhang, W. Li, R. Pei, F. Song, X. Liu, L. Kong, and Y. Zhang, “Dog-IQA: Standard-guided zero-shot mllm for mix-grained image quality assessment,”arXiv preprint arXiv:2410.02505, 2024

    Pith/arXiv arXiv 2024

  47. [47]

    Adaptive image quality assessment via teaching large multimodal model to compare,

    H. Zhu, H. Wu, Y. Li, Z. Zhang, B. Chen, L. Zhu, Y. Fang, G. Zhai, W. Lin, and S. Wang, “Adaptive image quality assessment via teaching large multimodal model to compare,”arXiv preprint arXiv:2405.19298, 2024

    Pith/arXiv arXiv 2024

  48. [48]

    Mitigating perception bias: A training-free approach to enhance lmm for image quality assessment,

    B. Chen, S. Pan, D. Wu, L. Xie, X. Sui, L. Zhu, and H. Zhu, “Mitigating perception bias: A training-free approach to enhance lmm for image quality assessment,”AAAI Conference on Artificial Intelligence, 2026. SUBMITTED TO IEEE TRANSACTIONS ON PATTERN ANAL YSIS AND MACHINE INTELLIGENCE 15

    2026

  49. [49]

    A perceptually weighted rank correlation indicator for objective image quality assessment,

    Q. Wu, H. Li, F. Meng, and K. N. Ngan, “A perceptually weighted rank correlation indicator for objective image quality assessment,” IEEE Transactions on Image Processing, vol. 27, no. 5, pp. 2499–2513, 2018

    2018

  50. [50]

    Maximum differentiation (mad) competition: A methodology for comparing computational mod- els of perceptual quantities,

    Z. Wang and E. P . Simoncelli, “Maximum differentiation (mad) competition: A methodology for comparing computational mod- els of perceptual quantities,”Journal of Vision, vol. 8, no. 12, pp. 8–8, 2008

    2008

  51. [51]

    Group maximum differentiation competition: Model comparison with few samples,

    K. Ma, Z. Duanmu, Z. Wang, Q. Wu, W. Liu, H. Yong, H. Li, and L. Zhang, “Group maximum differentiation competition: Model comparison with few samples,”IEEE Transactions on Pattern Anal- ysis and Machine Intelligence, vol. 42, no. 4, pp. 851–864, 2018

    2018

  52. [52]

    Waterloo exploration database: New challenges for image quality assessment models,

    K. Ma, Z. Duanmu, Q. Wu, Z. Wang, H. Yong, H. Li, and L. Zhang, “Waterloo exploration database: New challenges for image quality assessment models,”IEEE Transactions on Image Processing, vol. 26, no. 2, pp. 1004–1016, 2016

    2016

  53. [53]

    Eigen- distortions of hierarchical representations,

    A. Berardino, V . Laparra, J. Ball ´e, and E. Simoncelli, “Eigen- distortions of hierarchical representations,”Advances in neural information processing systems, vol. 30, 2017

    2017

  54. [54]

    Beta regression for modelling rates and proportions,

    S. Ferrari and F. Cribari-Neto, “Beta regression for modelling rates and proportions,”Journal of applied statistics, vol. 31, no. 7, pp. 799– 815, 2004

    2004

  55. [55]

    Uncertainty-aware blind image quality assessment in the laboratory and wild,

    W. Zhang, K. Ma, G. Zhai, and X. Yang, “Uncertainty-aware blind image quality assessment in the laboratory and wild,”IEEE Transactions on Image Processing, vol. 30, pp. 3474–3486, 2021

    2021

  56. [56]

    Descriptive image quality assessment in the wild,

    Z. You, J. Gu, Z. Li, X. Cai, K. Zhu, T. Xue, and C. Dong, “Descriptive image quality assessment in the wild,”arXiv preprint arXiv:2405.18842, 2024

    arXiv 2024

  57. [57]

    Delving into deep imbalanced regression,

    Y. Yang, K. Zha, Y. Chen, H. Wang, and D. Katabi, “Delving into deep imbalanced regression,” inInternational conference on machine learning. PMLR, 2021, pp. 11 842–11 851

    2021

  58. [58]

    KADID-10k: A large-scale ar- tificially distorted IQA database,

    H. Lin, V . Hosu, and D. Saupe, “KADID-10k: A large-scale ar- tificially distorted IQA database,” inInternational Conference on Quality of Multimedia Experience, 2019

    2019

  59. [59]

    PIPAL: a large-scale image quality assessment dataset for perceptual image restoration,

    G. Jinjin, C. Haoming, C. Haoyu, Y. Xiaoxing, J. S. Ren, and D. Chao, “PIPAL: a large-scale image quality assessment dataset for perceptual image restoration,” inEuropean Conf. on Computer Vision. Springer, 2020, pp. 633–651

    2020

  60. [60]

    Making a “com- pletely blind

    A. Mittal, R. Soundararajan, and A. C. Bovik, “Making a “com- pletely blind” image quality analyzer,”IEEE Signal Processing Letters, vol. 20, no. 3, pp. 209–212, 2013

    2013

  61. [61]

    Exploring CLIP for assessing the look and feel of images,

    J. Wang, K. C. Chan, and C. C. Loy, “Exploring CLIP for assessing the look and feel of images,” inAAAI Conference on Artificial Intelligence, 2023, pp. 1357–1365

    2023

  62. [62]

    Quality-aware image-text alignment for opinion-unaware image quality assessment,

    L. Agnolucci, L. Galteri, and M. Bertini, “Quality-aware image-text alignment for opinion-unaware image quality assessment,”arXiv preprint arXiv:2403.11176, 2024

    Pith/arXiv arXiv 2024

  63. [63]

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

    S. Yang, T. Wu, S. Shi, S. Lao, Y. Gong, M. Cao, J. Wang, and Y. Yang, “MANIQA: Multi-dimension attention network for no- reference image quality assessment,” inIEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022, pp. 1191–1200

    2022

  64. [64]

    KonIQ-10k: An eco- logically valid database for deep learning of blind image quality assessment,

    V . Hosu, H. Lin, T. Sziranyi, and D. Saupe, “KonIQ-10k: An eco- logically valid database for deep learning of blind image quality assessment,”IEEE Transactions on Image Processing, vol. 29, pp. 4041–4056, 2020

    2020

  65. [65]

    Massive online crowdsourced study of subjective and objective picture quality,

    D. Ghadiyaram and A. C. Bovik, “Massive online crowdsourced study of subjective and objective picture quality,”IEEE Transac- tions on Image Processing, vol. 25, no. 1, pp. 372–387, 2015

    2015

  66. [66]

    Perceptual quality assessment of smartphone photography,

    Y. Fang, H. Zhu, Y. Zeng, K. Ma, and Z. Wang, “Perceptual quality assessment of smartphone photography,” inIEEE Conference on Computer Vision and Pattern Recognition, 2020, pp. 3677–3686. Baoliang Chen(Member, IEEE) received his B.E. degree in Electronic Information Science and Technology from Hefei University of Tech- nology, Hefei, China, in 2015, his M...

    2020

  67. [2017]

    He has been an Associate Editor of IEEE TRANSACTIONS ON IMAGE PROCESSING, IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY , IEEE TRANSACTIONS ON MULTIMEDIA, and IEEE SIGNAL PROCESSING LETTERS. He was a Distinguished Lecturer of Asia-Pacific Signal and Information Process- ing Association (APSIPA) from 2012 to 2013 and the IEEE Circuits and ...

    2012

This paper was first reviewed by deepseek-v4-flash on August 3, 2026.