(50-4) 14 * << * >> * Russian * English * Content * All Issues

Decomposition of Perceptual Scores in Multi-Component IQA: Pathways to Enhanced Model Interpretability and Efficiency
A.M. Trykin1, V.E. Turlapov1

1 Artificial Intelligence Research Center, Lobachevsky State University,
603022, Nizhny Novgorod, Russia, Gagarin Ave., 23

  Full text (PDF)

DOI: 10.18287/COJ2082

Article ID: 2082

Language: English

Abstract:
Traditional Image Quality Assessment (IQA) often aggregates human perception into a single score, overlooking its multidimensional nature, which is crucial for interpretable IQA modeling. This study presents a systematic cross-dataset analysis of how technical distortions or aesthetic components jointly shape overall quality perception. By evaluating seven large-scale datasets (SPAQ, KonIQ-10k, GISET, ICAA20K, AADB, EVA, and PARA), we investigate the relationships between human-annotated attribute-level features and Mean Opinion Scores (MOS) using Pairwise Correlation (PC), Principal Component Analysis (PCA), and cross-validated regression models. Our analysis uncovers four distinct empirical patterns in perceptual score modeling: (1) high attribute redundancy (SPAQ, GISET, EVA, PARA), where inter-attribute correlations exceed 0.7-0.9, achieving near-optimal predictive power (PLCC/SRCC > 0.96) and indicating that existing multi-attribute annotations often collapse into a single dimension, rendering complex architectural partitioning redundant for these specific tasks; (2) effective orthogonality in AADB, where aesthetic attributes show weak inter-correlations (r < 0.3) yet enable accurate regression (PLCC/SRCC approximately 0.90), demonstrating that orthogonal features contribute unique variance; (3) an extraction and complexity gap in KonIQ-10k, where structurally simple, automated attributes fail to align with actual subjective perception (PLCC/SRCC approximately 0.74); and (4) insufficient specificity in ICAA20K, where isolated perceptual color attributes (temperature, harmony) show weak correlation with MOS (PLCC approximately 0.34), proving inadequate without semantic context. Based on these findings, we propose actionable design principles for future IQA benchmarks, prioritizing semantically distinct attributes alongside annotation protocols designed to mitigate cognitive biases (e.g., the halo effect), thereby ensuring interpretable and robust IQA models.

Keywords:
image quality assessment, image quality aesthetic assessment, multidimensional image quality assessment, correlation analysis, principal component analysis, machine learning, regression.

Citation:
Trykin AM, Turlapov VE. Decomposition of Perceptual Scores in Multi-Component IQA: Pathways to Enhanced Model Interpretability and Efficiency. Computer Optics 2026; 50(4): 2082. DOI: 10.18287/COJ2082.

References:

  1. Zhai G, Min X. Perceptual image quality assessment: A survey. Science China Information Sciences 2020; 63(11): 211301. DOI: 10.1007/s11432-019-2757-1.
  2. Kastryulin S, Zakirov J, Pezzotti N, Dylov DV. Image quality assessment for magnetic resonance imaging. IEEE Access 2023; 11: 14154-14168. DOI: 10.1109/ACCESS.2023.3243466.
  3. Jamil S. Review of image quality assessment methods for compressed images. Journal of Imaging 2024; 10(5): 113. DOI: 10.3390/jimaging10050113.
  4. Soydaner D, Wagemans J. Unveiling the factors of aesthetic preferences with explainable AI. British Journal of Psychology 2026; 117(2): 444-478. DOI: 10.1111/bjop.12707.
  5. Fang Y, Zhu H, Zeng Y, Ma K, Wang Z. Perceptual quality assessment of smartphone photography. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. IEEE; 2020: 3674-3683. DOI: 10.1109/CVPR42600.2020.00373.
  6. Hosu V, Lin H, Sziranyi T, Saupe D. KonIQ-10k: An ecologically valid database for deep learning of blind image quality assessment. IEEE Transactions on Image Processing 2020; 29: 4041-4056. DOI: 10.1109/TIP.2020.2967829.
  7. Utke M, Zadtootaghaj S, Schmidt S, Bosse S, Möller S. NDNetGaming: Development of a no-reference deep CNN for gaming video quality prediction. Multimedia Tools and Applications 2022; 81: 3181-3203. DOI: 10.1007/s11042-020-09144-6.
  8. He S, Xiao Y, Ming A, Ma H. Prompt-guided image color aesthetics assessment: Models, datasets and benchmarks. Information Fusion 2025; 114: 102706. DOI: 10.1016/j.inffus.2024.102706.
  9. Kong S, Shen X, Lin Z, Mech R, Fowlkes C. Photo aesthetics ranking network with attributes and content adaptation. In: Leibe B, Matas J, Sebe N, Welling M, eds. Computer Vision - ECCV 2016. Lecture Notes in Computer Science. Vol 9912. Springer; 2016: 662-679. DOI: 10.1007/978-3-319-46448-0_40.
  10. Kang C, Valenzise G, Dufaux F. EVA: An explainable visual aesthetics dataset. In: Proceedings of the Joint Workshop on Aesthetic and Technical Quality Assessment of Multimedia and Media Analytics for Societal Trends. ACM; 2020: 5-13. DOI: 10.1145/3423268.3423590.
  11. Yang Y, Xu L, Li L, et al. Personalized image aesthetics assessment with rich attributes. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. IEEE; 2022: 19829-19837. DOI: 10.1109/CVPR52688.2022.01924.
  12. Freedman DA. Statistical Models: Theory and Practice. 2nd ed. Cambridge: Cambridge University Press; 2009. DOI: 10.1017/CBO9780511815867.
  13. Cortes C, Vapnik V. Support-vector networks. Machine Learning 1995; 20(3): 273-297. DOI: 10.1007/BF00994018.
  14. Awad M, Khanna R. Support vector regression. In: Efficient Learning Machines. New York: Apress; 2015: 67-80. DOI: 10.1007/978-1-4302-5990-9_4.
  15. Breiman L, Friedman JH, Olshen RA, Stone CJ. Classification and Regression Trees. New York: Routledge; 2017. DOI: 10.1201/9781315139470.
  16. Goldberger J, Hinton GE, Roweis S, Salakhutdinov RR. Neighbourhood components analysis. In: Saul LK, Weiss Y, Bottou L, eds. Advances in Neural Information Processing Systems 17. Cambridge, MA: MIT Press; 2005: 513-520.
  17. Rosenblatt F. The perceptron: A probabilistic model for information storage and organization in the brain. Psychological Review 1958; 65(6): 386-408. DOI: 10.1037/H0042519.
  18. Pedregosa F, Varoquaux G, Gramfort A, et al. Scikit-learn: Machine learning in Python. Journal of Machine Learning Research 2011; 12: 2825-2830.
  19. Akiba T, Sano S, Yanase T, Ohta T, Koyama M. Optuna: A next-generation hyperparameter optimization framework. In: Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. ACM; 2019: 2623-2631. DOI: 10.1145/3292500.3330701.
  20. Watanabe S. Tree-structured Parzen estimator: Understanding its algorithm components and their roles for better empirical performance. arXiv preprint arXiv:2304.11127; 2023.
  21. Thorndike EL. A constant error in psychological ratings. Journal of Applied Psychology 1920; 4(1): 25-29. DOI: 10.1037/H0071663.
  22. Ryu K, Yanaka H. What do vision-language models encode for personalized image aesthetics assessment? In: Findings of the Association for Computational Linguistics: ACL 2026. Association for Computational Linguistics; 2026: 34146-34167. DOI: 10.18653/v1/2026.findings-acl.1706.

151, Molodogvardeiskaya str., Samara, 443001, Russia; E-mail: journal@computeroptics.ru; Tel: +7 (846) 242-41-24 (Executive secretary), +7 (846) 332-56-22 (Issuing editor), Fax: +7 (846) 332-56-20