Abstract
Despite remarkable success has been achieved by convolutional neural networks (CNNs) in no-reference image quality assessment (NR-IQA), there still exist many challenges in improving the performance of IQA for authentically distorted images. An important factor is that the insufficient annotated data limits the training of high-capacity CNNs to accommodate diverse distortions, complicated semantic structures and high-variance quality scores of these images. To address this problem, this paper proposes a hierarchical curriculum learning (HCL) framework for NR-IQA. The main idea of the proposed framework is to leverage the external data to learn the prior knowledge about IQA widely and progressively. Specifically, as a closely-related task with NR-IQA, image restoration is used as the first curriculum to learn the image quality related knowledge (i.e., semantic and distortion information) on massive distorted-reference image pairs. Then multiple lightweight subnetworks are designed to learn human scoring rules on multiple available synthetic IQA datasets independently, and a cross-dataset quality assessment correlation (CQAC) module is proposed to fully explore the similarities and diversities of different scoring rules. Finally, the whole model is fine-tuned on the target authentic IQA dataset to fuse the learned knowledge and adapt to the target data distribution. Experimental results show that our model achieves state-of-the-art performance on multiple standard authentic IQA datasets. Moreover, the generalization of our model is fully validated by the cross-dataset evaluation and the gMAD competition. In addition, extensive analyses prove that the proposed HCL framework is effective in improving the performance of our model.
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs11263-023-01851-5/MediaObjects/11263_2023_1851_Fig1_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs11263-023-01851-5/MediaObjects/11263_2023_1851_Fig2_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs11263-023-01851-5/MediaObjects/11263_2023_1851_Fig3_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs11263-023-01851-5/MediaObjects/11263_2023_1851_Fig4_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs11263-023-01851-5/MediaObjects/11263_2023_1851_Fig5_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs11263-023-01851-5/MediaObjects/11263_2023_1851_Fig6_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs11263-023-01851-5/MediaObjects/11263_2023_1851_Fig7_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs11263-023-01851-5/MediaObjects/11263_2023_1851_Fig8_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs11263-023-01851-5/MediaObjects/11263_2023_1851_Fig9_HTML.png)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs11263-023-01851-5/MediaObjects/11263_2023_1851_Fig10_HTML.png)
Similar content being viewed by others
Data Availability
The data that support the findings of this study are available from the corresponding author upon reasonable request.
References
A, T. D., B, K. G., C, L. N., et al. (2018). Referenceless quality metric of multiply-distorted images based on structural degradation. Neurocomputing, 290, 185–195.
Bosse, S., Maniry, D., Müller, K. R., et al. (2017). Deep neural networks for no-reference and full-reference image quality assessment. IEEE Transactions on Image Processing (TIP), 27(1), 206–219.
Ciancio, A., da Silva, E. A., Said, A., et al. (2010). No-reference blur assessment of digital pictures based on multifeature classifiers. IEEE Transactions on Image Processing (TIP), 20(1), 64–75.
Deng, J., Dong, W., Socher, R., et al. (2009). Imagenet: A large-scale hierarchical image database. In Conference on computer vision and pattern recognition (CVPR) (pp. 248–255).
Fang, Y., Ma, K., Wang, Z., et al. (2015). No-reference quality assessment of contrast-distorted images based on natural scene statistics. IEEE Signal Processing Letters, 22(7), 838–842. https://doi.org/10.1109/LSP.2014.2372333
Fang, Y., Zhu, H., Zeng, Y., et al. (2020). Perceptual quality assessment of smartphone photography. In Conference on computer vision and pattern recognition (CVPR).
Ghadiyaram, D., & Bovik, A. C. (2015). Massive online crowdsourced study of subjective and objective picture quality. IEEE Transactions on Image Processing (TIP), 25(1), 372–387.
Ghadiyaram, D., & Bovik, A. C. (2017). Perceptual quality prediction on authentically distorted images using a bag of features approach. Journal of Vision, 17(1), 32–32.
Golestaneh, S. A., Dadsetan, S., & Kitani, K. M. (2022). No-reference image quality assessment via transformers, relative ranking, and self-consistency. In IEEE Winter conference on applications of computer vision (WACV) (pp. 3989–3999). https://doi.org/10.1109/WACV51458.2022.00404
He, K., Zhang, X., Ren, S., et al. (2016). Deep residual learning for image recognition. In Conference on computer vision and pattern recognition (CVPR) (pp. 770–778).
Hosu, V., Lin, H., Sziranyi, T., et al. (2020). Koniq-10k: An ecologically valid database for deep learning of blind image quality assessment. IEEE Transactions on Image Processing (TIP), 29, 4041–4056.
Kang, L., Ye, P., Li, Y., et al. (2014). Convolutional neural networks for no-reference image quality assessment. In Conference on computer vision and pattern recognition (CVPR) (pp. 1733–1740).
Kang, L., Ye, P., Li, Y., et al. (2015). Simultaneous estimation of image quality and distortion via multi-task convolutional neural networks. In International conference on image processing (ICIP), IEEE (pp. 2791–2795).
Ke, J., Wang, Q., Wang, Y., et al. (2021). Musiq: Multi-scale image quality transformer. In International conference on computer vision (ICCV) (pp. 5148–5157).
Kim, J., & Lee, S. (2016). Fully deep blind image quality predictor. IEEE Journal of Selected Topics in Signal Processing, 11(1), 206–220.
Kim, J., Nguyen, A. D., & Lee, S. (2018). Deep cnn-based blind image quality predictor. IEEE Transactions on Neural Networks and Learning Systems, 30(1), 11–24.
Krizhevsky, A., Sutskever, I., & Hinton, G. E. (2012). Imagenet classification with deep convolutional neural networks. Advances in Neural Information Processing Systems (NeurIPS), 25, 1097–1105.
Larson, E. C., & Chandler, D. M. (2010). Most apparent distortion: Full-reference image quality assessment and the role of strategy. Journal of Electronic Imaging, 19(1), 011006.
Li, D., Jiang, T., Lin, W., et al. (2018). Which has better visual quality: The clear blue sky or a blurry animal? IEEE Transactions on Multimedia (TMM), 21(5), 1221–1234.
Li, D., Jiang, T., & Jiang, M. (2020). Norm-in-norm loss with faster convergence and better performance for image quality assessment. In ACM International conference on multimedia (ACM MM) (pp. 789–797). ACM.
Li, D., Jiang, T., & Jiang, M. (2021). Unified quality assessment of in-the-wild videos with mixed datasets training. International Journal of Computer Vision (IJCV), 129, 1238–1257.
Li, L., **a, W., Lin, W., et al. (2017). No-reference and robust image sharpness evaluation based on multiscale spatial and spectral features. IEEE Transactions on Multimedia (TMM), 19(5), 1030–1040. https://doi.org/10.1109/TMM.2016.2640762
Li, L., Song, T., Wu, J., et al. (2022). Blind image quality index for authentic distortions with local and global deep feature aggregation. IEEE Transactions on Circuits and Systems for Video Technology (TCSTV), 32(12), 8512–8523. https://doi.org/10.1109/TCSVT.2021.3112197
Li, Q., Lin, W., Xu, J., et al. (2016). Blind image quality assessment using statistical structural and luminance features. IEEE Transactions on Multimedia (TMM), 18(12), 2457–2469.
Li, X. (2002). Blind image quality assessment. In International conference on image processing (ICIP) (pp. I–I). https://doi.org/10.1109/ICIP.2002.1038057
Lin, H., Hosu, V., & Saupe, D. (2019). Kadid-10k: A large-scale artificially distorted iqa database. In Conference on quality of multimedia experience (QoMEX), IEEE (pp. 1–3).
Lin, K. Y., & Wang, G. (2018) Hallucinated-iqa: No-reference image quality assessment via adversarial learning. In Conference on computer vision and pattern recognition (CVPR) (pp. 732–741).
Liu, L., Dong, H., Huang, H., et al. (2014). No-reference image quality assessment in curvelet domain. Signal Processing: Image Communication, 29(4), 494–505.
Liu, X., Van De Weijer, J., & Bagdanov, A. D. (2017). Rankiqa: Learning from rankings for no-reference image quality assessment. In International conference on computer vision (ICCV) (pp. 1040–1049).
Ma, J., Wu, J., Li, L., et al. (2021). Blind image quality assessment with active inference. IEEE Transactions on Image Processing (TIP), 30, 3650–3663.
Ma, K., Duanmu, Z., Wu, Q., et al. (2016). Waterloo exploration database: New challenges for image quality assessment models. IEEE Transactions on Image Processing (TIP), 26(2), 1004–1016.
Ma, K., Liu, W., Liu, T., et al. (2017). dipiq: Blind image quality assessment by learning-to-rank discriminable image pairs. IEEE Transactions on Image Processing (TIP), 26(8), 3951–3964.
Ma, K., Duanmu, Z., Zhou, W., et al. (2018). Group maximum differentiation competition: Model comparison with few samples. IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), PP(99), 1–1.
Mittal, A., Moorthy, A. K., & Bovik, A. C. (2012). No-reference image quality assessment in the spatial domain. IEEE Transactions on Image Processing (TIP), 21(12), 4695–4708.
Moorthy, A. K., & Bovik, A. C. (2010). A two-step framework for constructing blind image quality indices. IEEE Signal Processing Letters, 17(5), 513–516. https://doi.org/10.1109/LSP.2010.2043888
Pan, X., Zhan, X., Dai, B., et al. (2022). Exploiting deep generative prior for versatile image restoration and manipulation. IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), 44(11), 7474–7489. https://doi.org/10.1109/TPAMI.2021.3115428
Pan, Z., Yuan, F., Lei, J., et al. (2022). Vcrnet: Visual compensation restoration network for no-reference image quality assessment. IEEE Transactions on Image Processing (TIP), 31, 1613–1627. https://doi.org/10.1109/TIP.2022.3144892
Ponomarenko, N., Ieremeiev, O., Lukin, V., et al. (2013). Color image database tid2013: Peculiarities and preliminary results. In European workshop on visual information processing (EUVIP) (pp. 106–111)
Ren, H., Chen, D., & Wang, Y. (2017). Ran4iqa: Restorative adversarial nets for no-reference image quality assessment. In AAAI conference on artificial intelligence (AAAI) (Vol. 32). https://doi.org/10.1609/aaai.v32i1.12258
Ren, H., Chen, D., & Wang, Y. (2018). Ran4iqa: Restorative adversarial nets for no-reference image quality assessment. In AAAI Conference on artificial intelligence (AAAI).
Saad, M. A., Bovik, A. C., & Charrier, C. (2012). Blind image quality assessment: A natural scene statistics approach in the dct domain. IEEE Transactions on Image Processing (TIP), 21(8), 3339–3352.
Sheikh, H. R., Sabir, M. F., & Bovik, A. C. (2006). A statistical evaluation of recent full reference image quality assessment algorithms. IEEE Transactions on Image Processing (TIP), 15(11), 3440–3451.
Simonyan, K., & Zisserman, A. (2014). Very deep convolutional networks for large-scale image recognition. ar**v preprint ar**v:1409.1556
Song, T., Li, L., Zhu, H., et al. (2021). Ie-iqa: Intelligibility enriched generalizable no-reference image quality assessment. Frontiers in Neuroscience, 15, 739138.
Song, T., Li, L., Chen, P., et al. (2022). Blind image quality assessment for authentic distortions by intermediary enhancement and iterative training. IEEE Transactions on Circuits and Systems for Video Technology (TCSVT), 32(11), 7592–7604. https://doi.org/10.1109/TCSVT.2022.3179744
Song, T., Li, L., Wu, J., et al. (2022). Knowledge-guided blind image quality assessment with few training samples. IEEE Transactions on Multimedia (TMM). https://doi.org/10.1109/TMM.2022.3233244
Su, S., Yan, Q., Zhu, Y., et al. (2020). Blindly assess image quality in the wild guided by a self-adaptive hyper network. In Conference on computer vision and pattern recognition (CVPR).
Tan, M., & Le, Q. (2019). Efficientnet: Rethinking model scaling for convolutional neural networks. In International conference on machine learning (ICML) (pp. 6105–6114).
Thomee, B., Shamma, D. A., Friedland, G., et al. (2016). Yfcc100m: The new data in multimedia research. Communications of the ACM, 59(2), 64–73.
Touvron, H., Cord, M., Douze, M., et al. (2021). Training data-efficient image transformers & distillation through attention. In International conference on machine learning (ICML) (pp. 10,347–10,357)
Vaswani, A., Shazeer, N., Parmar, N., et al. (2017). Attention is all you need. Advances in Neural Information Processing Systems, 30. https://proceedings.neurips.cc/paper_files/paper/2017/file/3f5ee243547dee91fbd053c1c4a845aa-Paper.pdf
Wang, J., Fan, H., Hou, X., et al. (2022). Mstriq: No reference image quality assessment based on swin transformer with multi-stage fusion. In Conference on computer vision and pattern recognition workshops (CVPRW) (pp. 1268–1277). https://doi.org/10.1109/CVPRW56347.2022.00133
Wang, Z., & Ma, K. (2021). Active fine-tuning from gmad examples improves blind image quality assessment. IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI).
Wang, Z., Bovik, A. C., Sheikh, H. R., et al. (2004). Image quality assessment: From error visibility to structural similarity. IEEE Transactions on Image Processing (TIP), 13(4), 600–612.
Wu, Q., Wang, Z., & Li, H. (2015). A highly efficient method for blind image quality assessment. In International conference on image processing (ICIP) (pp. 339–343). IEEE.
Xu, J., Ye, P., Li, Q., et al. (2016). Blind image quality assessment based on high order statistics aggregation. IEEE Transactions on Image Processing (TIP), 25(9), 4444–4457.
Xue, W., Mou, X., Zhang, L., et al. (2014). Blind image quality assessment using joint statistics of gradient magnitude and laplacian features. IEEE Transactions on Image Processing (TIP), 23(11), 4850–4862. https://doi.org/10.1109/TIP.2014.2355716
Yan, B., Bare, B., & Tan, W. (2019). Naturalness-aware deep no-reference image quality assessment. IEEE Transactions on Multimedia (TMM), 21(10), 2603–2615. https://doi.org/10.1109/TMM.2019.2904879
Yan, Q., Gong, D., & Zhang, Y. (2019). Two-stream convolutional networks for blind image quality assessment. IEEE Transactions on Image Processing (TIP), 28(5), 2200–2211. https://doi.org/10.1109/TIP.2018.2883741
Yan, Q., Gong, D., & Zhang, Y. (2019). Two-stream convolutional networks for blind image quality assessment. IEEE Transactions on Image Processing (TIP), 28(5), 2200–2211. https://doi.org/10.1109/TIP.2018.2883741
Yang, S., Wu, T., Shi, S., et al (2022). Maniqa: Multi-dimension attention network for no-reference image quality assessment. In Conference on computer vision and pattern recognition workshops (CVPRW) (pp. 1190–1199). https://doi.org/10.1109/CVPRW56347.2022.00126
Ye, P., Kumar, J., Kang, L., et al. (2012). Unsupervised feature learning framework for no-reference image quality assessment. In Conference on computer vision and pattern recognition (CVPR) (pp. 1098–1105). https://doi.org/10.1109/CVPR.2012.6247789
Yin, G., Wang, W., Yuan, Z., et al. (2022). Content-variant reference image quality assessment via knowledge distillation. In AAAI Conference on artificial intelligence (AAAI).
You, J., & Korhonen, J. (2021). Transformer for image quality assessment. In International conference on image processing (ICIP) (pp. 1389–1393). https://doi.org/10.1109/ICIP42928.2021.9506075
Zeng, H., Zhang, L., & Bovik, A. C. (2017). A probabilistic quality representation approach to deep blind image quality prediction. ar**v preprint ar**v:1708.08190
Zhang, L., Zhang, L., & Bovik, A. C. (2015). A feature-enriched completely blind image quality evaluator. IEEE Transactions on Image Processing (TIP), 24(8), 2579–2591.
Zhang, W., Ma, K., Yan, J., et al. (2018). Blind image quality assessment using a deep bilinear convolutional neural network. IEEE Transactions on Circuits and Systems for Video Technology (TCSVT), 30(1), 36–47.
Zhang, W., Ma, K., Zhai, G., et al. (2021). Uncertainty-aware blind image quality assessment in the laboratory and wild. IEEE Transactions on Image Processing (TIP), 30, 3474–3486.
Zhang, W., Li, D., Ma, C., et al. (2023). Continual learning for blind image quality assessment. IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), 45(3), 2864–2878. https://doi.org/10.1109/TPAMI.2022.3178874
Zhao, K., Yuan, K., Sun, M., et al. (2023). Quality-aware pre-trained models for blind image quality assessment. In Conference on computer vision and pattern recognition (CVPR).
Zhu, H., Li, L., Wu, J., et al. (2020). Metaiqa: Deep meta-learning for no-reference image quality assessment. In Conference on computer vision and pattern recognition (pp. 14,143–14,152).
Zhu, Y., Li, Y., Sun, W., et al. (2022). Blind image quality assessment via cross-view consistency. IEEE Transactions on Multimedia (TMM) 1–14. https://doi.org/10.1109/TMM.2022.3224319
Funding
This work is supported by the National Key Research and Development Program of China (Grant No. 2020AAA0106800), the Natural Science Foundation of China (Grant No. 62202470, 61972397, 62122086, U1936204, 62036011, 62192782, 61721004, U2033210), Bei**g Natural Science Foundation (Grant No. 4224093, JQ21017, L223003), the Major Projects of Guangdong Education Department for Foundation Research and Applied Research (Grant No. 2017KZDXM081, 2018KZDXM066), Guangdong Provincial University Innovation Team Project (Project No. 2020KCXTD045) and Youth Innovation Promotion Association, CAS.
Author information
Authors and Affiliations
Contributions
Juan Wang designed the framework, performed the research and wrote the paper. Zewen Chen conducted the experiments and analyzed the results. Chunfeng Yuan reorganized the structure of the paper and carried out the experimental analysis. Bing Li refined the idea of the paper and provided some technical guidance (e.g., network design and training approach). Wentao Ma gave useful technical comments (e.g., experience in image quality assessment). Weiming Hu reviewed the paper and examined the technique details (e.g., mathematical formula and algorithm description). All authors read and approved the final manuscript.
Corresponding author
Ethics declarations
Conflict of interest
The authors have no competing interests to declare that are relevant to the content of this article.
Additional information
Communicated by Stephen Lin.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary Information
Below is the link to the electronic supplementary material.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Wang, J., Chen, Z., Yuan, C. et al. Hierarchical Curriculum Learning for No-Reference Image Quality Assessment. Int J Comput Vis 131, 3074–3093 (2023). https://doi.org/10.1007/s11263-023-01851-5
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11263-023-01851-5