Multiscale structural similarity for image quality assessment
The DOI matches its Crossref record; 16 of the 16 references it cites were found in a registry or an index; none of the 14 sources its claims cite is checked for support yet.
Abstract
The structural similarity image quality paradigm is based on the assumption that the human visual system is highly adapted for extracting structural information from the scene, and therefore a measure of structural similarity can provide a good approximation to perceived image quality. This paper proposes a multiscale structural similarity method, which supplies more flexibility than previous single-scale methods in incorporating the variations of viewing conditions. We develop an image synthesis method to calibrate the parameters that define the relative importance of different scales. Experimental comparisons demonstrate the effectiveness of the proposed method.
Overview
CoChat Summary
TL;DR
This paper proposes a multi-scale structural similarity (MS-SSIM) image quality method that combines contrast and structure comparisons across scales and luminance comparison at the coarsest scale. On the LIVE JPEG/JPEG2000 database, MS-SSIM outperformed PSNR, the Sarnoff model, and single-scale SSIM variants on all reported evaluation criteria.
Key findings
- MS-SSIM achieved the best reported performance: CC 0.969, ROCC 0.966, MAE 3.86, RMS 4.91, and OR 1.16%, compared with the best single-scale SSIM at M=2 (CC 0.963, ROCC 0.959, MAE 4.21, RMS 5.38, OR 2.62%) and PSNR (CC 0.905, ROCC 0.901, MAE 6.53, RMS 8.45, OR 15.7%).
- Single-scale SSIM performance varied with scale; the best single-scale result was at M=2, and single-scale SSIM tended to give higher scores as scale increased.
- The calibrated cross-scale exponents were β1=γ1=0.0448, β2=γ2=0.2856, β3=γ3=0.3001, β4=γ4=0.2363, and α5=β5=γ5=0.1333.
- The authors conclude that multi-scale SSIM provides a meaningful balance between scales and outperforms the best single-scale SSIM and state-of-the-art image quality metrics.
Methods
- The MS-SSIM system iteratively applies a low-pass filter and downsamples by 2; at each scale it computes contrast and structure comparisons, computes luminance comparison only at Scale M, and combines them using exponents normalized so that ∑γj=1.
- Cross-scale parameters were calibrated by image synthesis: 5 scales and 12 distortion levels (range 2^3 to 2^14) were used with 10 original 64×64 images, producing 600 distorted images; 8 subjects selected images of equal perceived quality across scales at a viewing distance of 32 pixels per degree of visual angle.
- Evaluation used the LIVE database with 344 JPEG and JPEG2000 compressed images (bit rate 0.028 to 3.150 bits/pixel), MOS from 13–25 subjective scores, and compared PSNR, Sarnoff JNDmetrix 8.0, single-scale SSIM for M=1–5, and MS-SSIM using CC, MAE, RMS, OR, and ROCC after logistic mapping.
Limitations
- Stated: the image-synthesis calibration approach is “still rather crude,” and the authors are working on a more systematic approach.
- Stated: CSF cannot be directly used to derive the parameters because it is typically measured at visibility threshold with simplified stimuli, whereas the goal is comparing complex structured images at visible distortion levels.
- Evident from the text: calibration used only 10 original 64×64 grayscale images and 8 subjects, including one author, and evaluation was limited to the LIVE JPEG/JPEG2000 database.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
0 Valid · 1 Not valid · 14 Not checked · 18 Own results and interpretations
A result the paper states — it cites no source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“It can also be observed that the single-scale model tends to supply higher scores with the increase of scales.”
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Finally, for every one of the objective evaluation criteria, multi-scale SSIM model outperforms all the other models, including the best single-scale SSIM model, suggesting a meaningful balance between scales.”
- Also stated in: § 5. DISCUSSIONS · p. 4 · hedged
“Experiments show that with an appropriate parameter settings, the multi-scale method outperforms the best single-scale SSIM model as well as state-of-the-art image quality metrics.”
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Also stated in: § 5. DISCUSSIONS · p. 4 · hedged
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“In general, the subjects agreed with each other on each image more than they agreed with themselves across different images.”
plainstates an associationn = 8- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“The resulting parameters we obtained are β 1 = γ 1 = 0.0448, β 2 = γ 2 = 0.2856, β 3 = γ 3 = 0.3001, β 4 = γ 4 = 0.2363, and α 5 = β 5 = γ 5 = 0.1333, respectively.”
plainn = 8- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
In the paper
“However, the SSIM index algorithm introduced in [5] is a single-scale approach.”
What the sources say
- Not matchedcites [5] [5] Z. Wang, A. C. Bovik, H. R. Sheikh, and E. P. Simoncelli, 'Image quality assessment: From error measurement to structural similarity,' IEEE Trans. Image Processing , vol. 13, Jan. 2004. -
More than one paper fits this citation.
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Not matched
10 more claims
Create a free account to see all 33 claims and every source verdict.
- It may state two separate claims as one. 8
- Probably not a substantial claim. 13
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The cited source says what the claim attributes to it.
In the citing paper
“Further, as research in quality assessment has demonstrated, incorporating multiscale information when assessing quality produces QA algorithms that perform better in terms of correlation with human perception [13], [29].”
From No-Reference Image Quality Assessment in the Spatial Domain (2012) · cites this paper as [13]
Citations
Cites
Image quality assessment: from error visibility to structural similarity
Wang, Z., Bovik, A. C., Sheikh, H. R., & Simoncelli, E. P. (2004) · IEEE Transactions on Image Processing · doi:10.1109/tip.2003.819861 · Cited by 57,892
Self-citationA universal image quality index
Wang, Z., & Bovik, A. C. (2002) · IEEE Signal Processing Letters · doi:10.1109/97.995823 · Cited by 5,757
Self-citationImage quality measures and their performance
Eskicioglu, A. M., & Fisher, P. S. (1995) · IEEE Transactions on Communications · doi:10.1109/26.477498 · Cited by 2,117 · cited 1×
“Traditional perceptual image quality assessment methods are based on a bottom-up approach which attempts to simulate the functionality of the relevant early human visual system (HVS) components.”
BackgroundA Parametric Texture Model Based on Joint Statistics of Complex Wavelet Coefficients
Portilla, J., & Simoncelli, E. P. (2000) · International Journal of Computer Vision · doi:10.1023/a:1026553619983 · Cited by 1,806 · cited 2×
“These synthesis methods provide a powerful and efficient means of testing a model, and have the added benefit that the resulting images suggest improvements that might be made to the model [11].”
MethodSelf-citation(1995) · Medical Entomology and Zoology · Cited by 1,212
Pyramid-based texture analysis/synthesis
Heeger, D. J., & Bergen, J. R. (1995) · Proceedings of the 22nd annual conference on Computer graphics and interactive techniques - SIGGRAPH '95 · doi:10.1145/218380.218446 · Cited by 1,095
Video quality assessment based on structural distortion measurement
Wang, Z., Lu, L., & Bovik, A. C. (2004) · Signal Processing: Image Communication · doi:10.1016/s0923-5965(03)00076-6 · Cited by 1,024
Self-citationImage Quality Assessment: From Error Measurement to Structural Similarity
Wang, Z., Bovik, A. C., Sheikh, H. R., & Simoncelli, E. P. (2004) · Cited by 919
Self-citationTeo, P. C., & Heeger, D. J. (2002) · Proceedings of 1st International Conference on Image Processing · doi:10.1109/icip.1994.413502 · Cited by 471
Objective video quality assessment
(2003) · Handbook of Video Databases · doi:10.1201/9780203489864-45 · Cited by 359
Research works printed but not linked: 2
- 0, JNDmetrix Technology, Sarnoff Corporation
- 2000, Final report from the video quality experts group on the validation of objective models of video quality assessment, VQEG
Cited by
Items 3,001–3,100 of 6,047
Modified Pixels Based Fast Median Filter in Impulse Noise Environments
Lee, M. H., & Shin, S. Y. (2018) · Indonesian Journal of Electrical Engineering and Computer Science · doi:10.11591/ijeecs.v9.i3.pp731-741 · Cited by 6
Low complexity intra prediction mode decision for 3D-HEVC depth coding
Jaballah, S., Larabi, M., & Tahar, J. B. (2018) · Signal Processing Image Communication · doi:10.1016/j.image.2018.05.007 · Cited by 6
Enhancing Temporal Quality Measurements in a Globally Deployed Streaming Video Quality Predictor
Bampis, C. G., Li, Z., & Bovik, A. C. (2018) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2018.8451275 · Cited by 5
“As expected, the multiscale information captured by MS-SSIM yielded further performance gains, but most of these spatial metrics were outperformed by video-based methods, which demonstrates the importance of capturing temporal information.”
BackgroundInfluentialSelf-citationAfonin, V. V., Erofeeva, I. V., Zotkina, M. M., et al. (2019) · Vestnik MGSU · doi:10.22227/1997-0935.2019.1.83-93 · Cited by 6
Embedding Data Crypted With Extended Shifting Polybius Square Supporting Turkish Character Set
MACİT, H. B., Koyun, A., & Yüksel, M. E. (2019) · Bitlis Eren Üniversitesi Fen Bilimleri Dergisi · doi:10.17798/bitlisfen.455126 · Cited by 6
Cross-language Speech Dependent Lip-synchronization
Jha, A., Voleti, V., Namboodiri, V. P., & Jawahar, C. V. (2019) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp.2019.8682275 · Cited by 10
“Quality of generation: We also compare SSIM index [27], a widely used metric to evaluate the quality of generated images.”
MethodCompressed Image Quality Assessment Based on Saak Features
Zhang, X., Kwong, S., & Kuo, C. ‐. J. (2019) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2019.8803184 · Cited by 6
“In the past decades, to obtain more consistent quality measures with human visual perception, numerous image quality assessment (IQA) methods [4, 5, 6, 7, 8, 9, 10] and datasets [11, 12, 13] have been proposed for different distortion types.”
BackgroundMethodInfluentialNo-reference stereoscopic image quality assessment based on cyclopean image and enhanced image
Li, S., Ding, Y., & Chang, Y. (2019) · Signal Image and Video Processing · doi:10.1007/s11760-019-01582-6 · Cited by 6
“Images show different characteristics at different scales [26].”
BackgroundD’Souza, M., van Munster, C. E. P., Dorn, J. F., Dorier, A., Kamm, C. P., Steinheimer, S., Dahlke, F., Uitdehaag, B. M. J., Kappos, L., & Johnson, M. (2020) · Journal of Medical Internet Research · doi:10.2196/16669 · Cited by 9
“The network was trained using a multi-scale structural similarity–based perceptual loss function [13] with Kullback-Leibler regularization as per Kingma and Welling [10].”
MethodDeep Learning-Based Adaptive Image Compression System for a Real-World Scenario
Anelli, V. W., Deldjoo, Y., Noia, T. D., et al. (2020) · IEEE Conference on Evolving and Adaptive Intelligent Systems (EAIS) · doi:10.1109/eais48028.2020.9122753 · Cited by 7
“Consequently, some works tend to replace it with a more complex weighted combination of different visual quality metrics (e.g., MS-SSIM [22], see the work described in [15]) or different types of losses, e.g., perceptual [23] and adversarial [24] loss (see the works presented in [12], [20]).”
MethodData Augmentation Powered by Generative Adversarial Networks
Póka, K. B., & Szemenyei, M. (2020) · doi:10.1109/ismcr51255.2020.9263725 · Cited by 6
“The loss (could be MSE, SSIM [10] etc.) is computed between the given and the generated images [12].”
BackgroundA Variational Auto-Encoder Approach for Image Transmission in Noisy Channel
Estiri, A. H., Sabramooz, M. R., Banaei, A., Dehghan, A. H., Jamialahmadi, B., & Siavoshani, M. J. (2020) · International Symposium on Telecommunications · doi:10.1109/ist50524.2020.9345813 · Cited by 13
Autoencoder Based Inter-Vehicle Generalization for In-Cabin Occupant Classification
Cruz, S., Taetz, B., Wasenmüller, O., Stifter, T., & Stricker, D. (2021) · IEEE Intelligent Vehicles Symposium · doi:10.1109/iv48863.2021.9575641 · Cited by 6
“In this work, we consider for the reconstruction loss the mean squared error (MSE): r(a, b) = ∥a − b∥22, the structural similarity index (SSIM) [37]: r(a, b) = 1 − SSIM(a, b), the multiscale structural similarity index (MS-SSIM) [38]: r(a, b) = 1 −MSSSIM(a, b) and the perceptual (PC) loss [39]: sum of MSE errors after…”
InfluentialAdversarial Data Augmentation on Breast MRI Segmentation
Teixeira, J. F., Dias, M., Batista, E., Cósta, J., Teixeira, L. F., & Oliveira, H. P. (2021) · Applied Sciences · doi:10.3390/app11104554 · Cited by 8
“The SSIM [19] is a three-aspect metric that looks at the luminance distortion (l), contrast distortion (c) and loss of structural correlation (s).”
BackgroundMethodHuang, Z., Lin, K., Jia, C., Wang, S., & Ma, S. (2021) · IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw53098.2021.00207 · Cited by 6
“Compared to MSE, multi-scale structure similarity (MS-SSIM) index [9] has higher relationship with the subjective quality, and it is also deployed as a widely accepted distortion measurement to reflect the degree of human visual satisfaction.”
BackgroundInfrared and visible image fusion based on Laplacian pyramid and generative adversarial network
Wang, J., Ke, C., Wu, M., Liu, M., & Zeng, C. (2021) · KSII Transactions on Internet and Information Systems · doi:10.3837/tiis.2021.05.010 · Cited by 7
“mutual information (MI) [28], average gradient (AG) [29], edge intensity (EI) [30], multi-scale structural similarity (MS_SSIM) [31], quality of edge ( abf Q ) [32], and the sum of the correlations of differences (SCD) [33].”
BackgroundCompression of Time Evolutionary Image Data through Predictive Deep Neural Networks
Roy, R., Sato, K., Bhattachrya, S., et al. (2021) · IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing (CCGrid) · doi:10.1109/ccgrid51090.2021.00014 · Cited by 6
Deep Graphics Encoder for Real-Time Video Makeup Synthesis from Example
Kips, R., Jiang, R., Ba, S., et al. (2021) · IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw53098.2021.00431 · Cited by 6
Enhanced Tone Mapping Using Regional Fused GAN Training with a Gamma-Shift Dataset
Jung, S., Kwon, H., & Lee, S.-H. (2021) · Applied Sciences · doi:10.3390/app11167754 · Cited by 6
Green, D. S. S., Gelb, A., & Luke, G. P. (2021) · Journal of Imaging · doi:10.3390/jimaging7100201 · Cited by 6
“The MS-SSIM index incorporates image details at different resolutions to provide an image quality assessment based on the human visual system [29].”
MethodPUGCQ: A Large Scale Dataset for Quality Assessment of Professional User-Generated Content
Guo, L., Chen, B., Zhu, L., He, Q., Fan, H., & Wang, S. (2021) · ACM International Conference on Multimedia (ACM MM) · doi:10.1145/3474085.3475183 · Cited by 10
“Typically, conventional full-reference (FR) based VQA methods rely on the signal level comparison, and the popular IQA methods can be adopted as the backbone, including peak signal-to-noise ratio (PSNR), structural similarity index (SSIM) [44], multiscale structural similarity [45], visual signal-to-noise ratio (VSNR)…”
MethodGet The Best of the Three Worlds
Zheng, Z., Wang, X., Lin, X., et al. (2021) · ACM International Conference on Multimedia (ACM MM) · doi:10.1145/3474085.3475667 · Cited by 6
“For FasiNet-H, which is suitable for scenarios with strict requirements on distortion, we set the MS-SSIM indicator to 0.997.”
MethodInfluentialA New Edge and Pixel-Based Image Quality Assessment Metric for Colour and Depth Images
Mousavi, S. M. H., & Mosavi, S. M. H. (2022) · Iranian Joint Congress on Fuzzy and Intelligent Systems · doi:10.1109/cfis54774.2022.9756490 · Cited by 7
Identity Preserving Loss for Learned Image Compression
Xiao, J., Aggarwal, L., Banerjee, P., Aggarwal, M., & Medioni, G. G. (2022) · IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw56347.2022.00067 · Cited by 10
“We evaluate image quality in terms of Peak signal-to-noise ratio (PSNR) and Multi-scale structural similarity index measure (MS-SSIM) [33] metrics.”
BackgroundMethodInfluentialPrivacy-Preserving Image Template Sharing Using Contrastive Learning
Rezaeifar, S., Voloshynovskiy, S., Jirhandeh, M. A., & Kinakh, V. (2022) · Entropy · doi:10.3390/e24050643 · Cited by 8
“The average SSIM and MSSIM for reconstructed images from the protected features and three other scenarios are reported in Table 2.”
MethodA Novel Model-Based Defogging Method for Particle Images With Different Fog Distributions
Zhou, S., Liu, X., Duan, J., & Herz, F. (2022) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2022.3164154 · Cited by 6
A 4-D Ultrasound Tomography for Industrial Process Reactors Investigation
Koulountzios, P., Rymarczyk, T., & Soleimani, M. (2022) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2022.3164166 · Cited by 6
“A more advanced form of SSIM, called Multiscale SSIM (MS-SSIM), is conducted over multiple scales through a process of multiple stages of sub-sampling [46].”
MethodVideo Coding using Learned Latent GAN Compression
Shukor, M., Damodaran, B. B., Yao, X., & Pierre Hellier, J. (2022) · ACM International Conference on Multimedia (ACM MM) · doi:10.1145/3503161.3548219 · Cited by 11
Low-Light Color Imaging via Cross-Camera Synthesis
Guo, P., Asif, M. S., & Ma, Z. (2022) · IEEE Journal of Selected Topics in Signal Processing · doi:10.1109/jstsp.2022.3175015 · Cited by 6
Datta, S., Nowrouzezahrai, D., Schied, C., & Dong, Z. (2022) · ACM SIGGRAPH Conference (SIGGRAPH) · doi:10.1145/3528233.3530700 · Cited by 9
“Several techniques exist for measuring perceptual similarity with respect to a single reference frame [Andersson et al. 2020; Wang et al. 2003] or reference video [Mantiuk et al. 2021].”
MethodNo-Reference Quality Assessment of Transmitted Stereoscopic Videos Based on Human Visual System
Hasan, M. M., Islam, M. A., Rahman, S., et al. (2022) · Applied Sciences · doi:10.3390/app121910090 · Cited by 6
ADGNet: Attention Discrepancy Guided Deep Neural Network for Blind Image Quality Assessment
Ma, X., Wang, Y., Liu, C., Zhang, S., & Yu, D. (2022) · ACM International Conference on Multimedia (ACM MM) · doi:10.1145/3503161.3547929 · Cited by 7
“Multi-scale features have shown promising potential in performance enhancement for quality assessment models [7, 14, 30].”
BackgroundBediako, D. O., Mou, X., & Suobogbiree, M. (2022) · Signal Image and Video Processing · doi:10.1007/s11760-022-02372-3 · Cited by 7
Progressive image dehazing network based on dual feature extraction modules
Yang, Y., Hu, W., Huang, S., Wan, W., & Guan, J. (2022) · International Journal of Machine Learning and Cybernetics · doi:10.1007/s13042-022-01753-x · Cited by 6
“To better evaluate the structural similarity between the dehazed and groundtruth images at different scales, a multiscale SSIM (MS_SSIM) quality evaluation index [33] is utilized as one term of the joint loss function, and it is defined as follows:”
MethodDocument Image Rectification in Complex Scene Using Stacked Siamese Networks
Xu, Z., Yin, F., Yang, P., et al. (2022) · Proceedings - International Conference on Pattern Recognition/Proceedings/International Conference on Pattern Recognition · doi:10.1109/icpr56361.2022.9956331 · Cited by 6
“Table II shows the comparison of different methods using MS-SSIM and LD on the same benchmark dataset.”
MethodInfluentialQuality-Driven Dual-Branch Feature Integration Network for Video Salient Object Detection
Zhou, X., Gao, H., Yu, L., et al. (2023) · Electronics · doi:10.3390/electronics12030680 · Cited by 6
Multi-Level Feature Aggregation Network for Full-Reference Image Quality Assessment
Chen, Z., Chen, Y., Liang, X., & Tang, Z. (2022) · IEEE International Conference on Tools with Artificial Intelligence (ICTAI) · doi:10.1109/ictai56018.2022.00132 · Cited by 6
“[25] 0.364 0.302 UQI [26] 0.505 0.461 SSIM [18] 0.377 0.319 MS-SSIM [19] 0.119 0.338 RFSIM [27] 0.285 0.254 GSM [28] 0.450 0.379 SRSIM [29] 0.626 0.529 FSIM [8] 0.553 0.452 VSI [9] 0.493 0.411 NIQE [30] 0.129 0.012 Brisque [31] 0.052 0.008 LPIPS-Alex [10] 0.606 0.569 LPIPS-VGG [10] 0.611 0.551…”
InfluentialFusion2Fusion: An Infrared–Visible Image Fusion Algorithm for Surface Water Environments
Lu, C. Y., Qin, H., Deng, Z., & Zhu, Z. (2023) · Journal of Marine Science and Engineering · doi:10.3390/jmse11050902 · Cited by 6
Chen, Y., Sun, Y., Bie, C., et al. (2023) · Quantitative Imaging in Medicine and Surgery · doi:10.21037/qims-22-1379 · Cited by 6
Large-Scale and Efficient Texture Mapping Algorithm via Loopy Belief Propagation
Ling, X., & Qin, R. (2023) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2023.3274091 · Cited by 6
“Then, the difference between the rendered and the original image is assessed using the Peak signal-to-noise ratio (PSNR) and the multiscale structural similarity index measure (MS-SSIM) [32].”
MethodFull-Reference Image Quality Assessment via Low-Level and High-Level Feature Fusion
Wu, C., Liao, X., Yue, H., Xu, X., Wei, X., Wu, D. M., & Zhou, M. (2023) · International Journal of Pattern Recognition and Artificial Intelligence · doi:10.1142/s0218001423540162 · Cited by 6
BlendFields: Few-Shot Example-Driven Facial Modeling
Kania, K., Garbin, S. J., Tagliasacchi, A., et al. (2023) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52729.2023.00047 · Cited by 6
“We use Peak Signal To Noise Ratio (PSNR) [3], Structural Similarity Index (SSIM) [60] and Learned Perceptual Image Patch Similarity (LPIPS) [77] to measure the performance of the models.”
MethodTowards Thousands to One Reference: Can We Trust the Reference Image for Quality Assessment?
Tian, Y., Chen, B., Wang, S., & Kwong, S. (2023) · IEEE Transactions on Multimedia · doi:10.1109/tmm.2023.3310268 · Cited by 8
“Several classical and state-of-the-art FR-IQA models, including PNSR, SSIM [9], MS-SSIM [82], FSIM [38], VSI [83], VIF [40], NLPD [84], GMSD [39], MAD [76], Deep-IQA [41], PieAPP [42], LPIPS [43], DISTS [44] and Deep-WSD [45], are employed for performance comparison.”
MethodJafari, F., & Basu, A. (2023) · Sensors · doi:10.3390/s23187790 · Cited by 11
Enhanced Diagnostic Fidelity in Pathology Whole Slide Image Compression via Deep Learning
Fischer, M., Neher, P., Schüffler, P. J., Xiao, S., Almeida, S. D., Ulrich, C., Muckenhuber, A., Braren, R., GOETZ, M. A., Kleesiek, J., et al. (2023) · Lecture notes in computer science · doi:10.1007/978-3-031-45676-3_43 · Cited by 10
“For two windows u and u ′ , which are located in x and x ′ respectively, the MS-SSIM is calculated as [17]: with the averaged intensities in the given block window µ , the variance σ , the covariance σ u,u ′ and the parameters c 1 , c 2 for stabilizing the division by c 1 = ( k 1 L ) 2 and c 2 = (…”
InfluentialAttention-Aware Patch-Based CNN for Blind 360-Degree Image Quality Assessment
Sendjasni, A., & Larabi, M. (2023) · Sensors · doi:10.3390/s23218676 · Cited by 9
“The selected models include PSNR, SSIM, MS-SSIM [62], FSIM [63], BRISQUE [64], and BMPRI [65], DB-CNN [34], and DipIQ [66], representing 2D-IQA models.”
MethodVET: Vasculature Extraction Transformer for Single-Scan Optical Coherence Tomography Angiography
Liao, J., Zhang, T., Zhang, Y., Li, C., & Huang, Z. (2023) · IEEE Transactions on Biomedical Engineering · doi:10.1109/tbme.2023.3330681 · Cited by 6
“To conduct a quantitative performance comparison of various methods, including SV-OCTA, ED-OCTA, and deep-learning-based methods, this study utilized peak-signal-to-noise ratio (PSNR), structural similarity (SSIM) [44] and multi-scale (MS)- SSIM [45] as objective evaluation metrics.”
MethodQin, Y., Butola, A., & Agarwal, K. (2023) · Inverse Problems · doi:10.1088/1361-6420/ad0c29 · Cited by 6
Learning Diverse Tone Styles for Image Retouching
Wang, H., Zhang, J., Liu, M., Wu, X., & Zuo, W. (2023) · IEEE Transactions on Image Processing · doi:10.1109/tip.2023.3340522 · Cited by 13
“After training, we follow the previous works [7], [17], which calculate the PSNR [44], SSIM [45], MS-SSIM [46], LPIPS [47], and NIMA [48] metrics for the MIT-Adobe FiveK dataset [16], and compare the PSNR [44] and CIELAB color difference ( (cid:49) E ab ) for the PPR10K dataset [17].”
MethodEfficient Deep-Detector Image Quality Assessment Based on Knowledge Distillation
Lee, W., Gold, G. E., & Choi, J. (2023) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2023.3346519 · Cited by 7
“We selected eight FR-IQA methods: RMSE, PSNR, SSIM [5], MS-SSIM [21], gradient magnitude similarity deviation (GMSD) [59], FSIM [23], noise quality measure (NQM) [60], and VIF [22], and five NR-IQA methods: NIQE [25], BRISQUE [24], PIQE [26], NPWE [37], [38], and D2IQA [17].”
MethodCar-Studio: Learning Car Radiance Fields From Single-View and Unlimited In-the-Wild Images
Liu, T., Zhao, H., Yu, Y., Zhou, G., & Liu, M. (2024) · IEEE Robotics and Automation Letters · doi:10.1109/lra.2024.3349949 · Cited by 13
“To quantitatively evaluate the results, we employ three metrics: Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index (SSIM) [56], and Learned Perceptual Image Patch Similarity (LPIPS) [57].”
MethodLRENet: a location-related enhancement network for liver lesions in CT images
Guo, S., Wang, H., Agaian, S. С., Han, L., & Song, X. (2024) · Physics in Medicine and Biology · doi:10.1088/1361-6560/ad1d6b · Cited by 7
Quantitative Metrics for Benchmarking Medical Image Harmonization
Parida, A., Jiang, Z., Packer, R. J., et al. (2024) · IEEE International Symposium on Biomedical Imaging (ISBI) · doi:10.1109/isbi56570.2024.10635289 · Cited by 6
“Metrics like structural similarity (SSIM) [10] can be used to alleviate the constraint on paired data [2], but they primarily compare structures at a higher level and could miss smaller hallucinations generated by GANs, which are clinically relevant.”
MethodFang, J., Wang, X., Li, Y., et al. (2024) · Remote Sensing · doi:10.3390/rs16081450 · Cited by 6
“As evident from the fi gure, our dataset consists of authentic haze images, displaying signi fi cant variability in the dehazing e ff ectiveness of each method.”
Rethinking Regularization with Random Label Smoothing
dos Santos, C. F. G., & Papa, J. P. (2024) · Neural Processing Letters · doi:10.1007/s11063-024-11579-z · Cited by 7
“Table5shows the outcomes of PyNET using RLS algorithm considering PSNR and Multiscale Structural Similarity Index Measure (MS-SSIM) [25] quality measures.”
Multi‐style cartoonization: Leveraging multiple datasets with generative adversarial networks
Cai, J., Li, F. W. B., Nan, F., & Yang, B. (2024) · Computer Animation and Virtual Worlds · doi:10.1002/cav.2269 · Cited by 6
No-Reference Multi-Level Video Quality Assessment Metric for 3D-Synthesized Videos
Wang, G., Huang, B., Gu, K., Liu, Y., Liu, H., Shi, Q., Zhai, G., & Zhang, W. (2024) · IEEE Transactions on Broadcasting · doi:10.1109/tbc.2024.3396696 · Cited by 6
“Subsequently, many researchers have made significant improvements based on SSIM, such as Multi-Scale SSIM (MS-SSIM) [9], Information Content Weighted SSIM (IW-SSIM) [10], and Analysis of Distortion Distribution based SSIM (ADD-SSIM) [11], etc.”
BackgroundRamyashree, R., Venugopala, P. S., Raghavendra, S., & Sarojadevi (2024) · Journal of Engineering · doi:10.1155/2024/8860475 · Cited by 7
Compression techniques for IoT enabled handheld ultrasound imaging system
Akkala, V., Bharath, R., Rajalakshmi, P., et al. (2014) · IEEE-EMBS Conference on Biomedical Engineering and Sciences · doi:10.1109/iecbes.2014.7047584 · Cited by 6
Comparison of objective quality assessment methods for scalable video coding
Vranješ, D., Žagar, D., & Nemcic, O. (2012) · International Symposium ELMAR (ELMAR) · Cited by 7
HDR2014 - A high dynamic range image quality database
Liu, M., Zhai, G., Tan, S., Zhang, Z., Gu, K., & Yang, X. (2014) · IEEE International Conference on Multimedia and Expo workshops · doi:10.1109/icmew.2014.6890713 · Cited by 6
A novel NMF-based image quality assessment metric using extreme learning machine
Wang, S., Deng, C., Lin, W., et al. (2013) · IEEE China Summit and International Conference on Signal and Information Processing (ChinaSIP) · doi:10.1109/chinasip.2013.6625339 · Cited by 6
A study on using spectral saliency detection approaches for image quality assessment
Saha, A., & Wu, Q. M. J. (2013) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp.2013.6637981 · Cited by 5
Microarchitectural analysis of image quality assessment algorithms
Phan, T. D., Shah, S. K., Chandler, D. M., & Sohoni, S. (2014) · Journal of Electronic Imaging · doi:10.1117/1.jei.23.1.013030 · Cited by 10
“D matrix via a thin C++ wrapper class.”
MethodInfluentialNovel image quality metric based on similarity
Jin, L.-N., Ponomarenko, N., & Egiazarian, K. (2011) · International Symposium on Signals, Circuits and Systems · doi:10.1109/isscs.2011.5978673 · Cited by 9
Face detection on distorted images using perceptual quality-aware features
Gunasekar, S., Ghosh, J., & Bovik, A. C. (2014) · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · doi:10.1117/12.2037343 · Cited by 6
Self-citationMulti-scale structural image quality assessment based on two-stage low-level features
Guo, L., Chen, W.-L., Liao, Y., Liao, H., & Hu, J. C. (2014) · Computers & Electrical Engineering · doi:10.1016/j.compeleceng.2014.01.004 · Cited by 6
A reduced-reference perceptual quality metric for texture synthesis
Varadarajan, S., & Karam, L. J. (2014) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2014.7025106 · Cited by 6
“The SSIM [6] and the multi-scale SSIM (MSSSIM) [7] metrics, use the mean, variance, and co-variance of pixels to compute structural similarity.”
BackgroundMethodInfluentialA Study of FR Video Quality Assessment of Real Time Video Stream
Zhang, X., Wu, L., Fang, Y., & Jiang, H. (2012) · International Journal of Advanced Computer Science and Applications · doi:10.14569/ijacsa.2012.030601 · Cited by 7
Zero and infinity images in multi-scale image fusion
Hossny, M., Nahavandi, S., & Creighton, D. (2009) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2009.5413827 · Cited by 6
“The fusion has been performed at 6 levels of detail (derived by laplacian analysis) and the average quality distances (MS-SSIM) [13] Δlm,n(x⊕0,x) and Δlm,n(x⊕∞,∞) have been calculated and plotted.”
BackgroundMethodBalanced sampling and compression for remote visualization
Frey, S., Sadlo, F., & Ertl, T. (2015) · doi:10.1145/2818517.2818529 · Cited by 6
Model-based full reference image blurriness assessment
Khosravi, M. H., & Hassanpour, H. (2016) · Multimedia Tools and Applications · doi:10.1007/s11042-015-3149-5 · Cited by 6
“PCC SROCC KROCC RMSE As another validation step, we compare the results of the proposed method with the results of some other well accepted quality metrics including PSNR, SSIM [27], MS-SIM [26], ADM [13], VIF [17], MAD [12], FSIM [29], SVDR [14], VGS [30] and the SVD-ELM based method proposed in [24].”
ResultLossy compression of images corrupted by spatially correlated noise
Krivenko, S., Kolganova, O., & Лукин, В. В. (2016) · International Conference on Modern Problems of Radio Engineering, Telecommunications and Computer Science · doi:10.1109/tcset.2016.7452157 · Cited by 6
Yang, M., Liu, J., Li, Z., Liang, L., Wang, X., & Gui, Z. (2016) · Neurocomputing · doi:10.1016/j.neucom.2016.05.018 · Cited by 6
“The formula for calculation of the SSIM is [26,27]”
MethodVisually lossless compression of synthetic aperture radar images
Ponomarenko, N. N., Лукин, В. В., & Egiazarian, K. (2011) · doi:10.1109/icatt.2011.6170755 · Cited by 6
Sparsity based stereoscopic image quality assessment
Khan, S., & Channappayya, S. S. (2016) · Asilomar Conference on Signals, Systems and Computers (ACSSC) · doi:10.1109/acssc.2016.7869706 · Cited by 5
“With this structural motivation we use MSSSIM [8] on both luminance and disparity pairs and pooled the scores.”
MethodNo-reference perceived image quality measurement for multiple distortions
Lamb, A. B., & Khambete, M. (2017) · Multimedia Tools and Applications · doi:10.1007/s11042-017-4761-3 · Cited by 6
“The performance of the proposed metric was also compared with that of state-of-theart IQA metrics, including FR-IQA methods PSNR, structural similarity (SSIM) index [25], multiscale SSIM (MS-SSIM) index [27], visual information fidelity (VIF) index [21], most apparent distortion (MAD) index [9], information content we…”
MethodA New Class of Wavelet-Based Metrics for Image Similarity Assessment
Albanesi, M. G., Amadeo, R., Bertoluzza, S., & Maggi, G. (2017) · Journal of Mathematical Imaging and Vision · doi:10.1007/s10851-017-0745-1 · Cited by 6
“Let us then look more closely at the differences between WNRMSE and, respectively, MSSSIM and CWSSIM.”
MethodInfluentialMulti-exposure image fusion quality assessment using contrast information
Lu, X., Zeng, H., Chen, J., et al. (2017) · International Symposium on Intelligent Signal Processing and Communication Systems (ISPACS) · doi:10.1109/ispacs.2017.8265641 · Cited by 6
“Inspired by [16], the multi-scale method is further adopted to evaluate the image quality of MEF from difference scales.”
MethodDepth Assisted Adaptive Workload Balancing for Parallel View Synthesis
Jin, X., Liu, Z., Li, Q., et al. (2018) · IEEE Transactions on Multimedia · doi:10.1109/tmm.2018.2827781 · Cited by 6
“As shown in the tables, the overall differences among the algorithms are very small, i.e., the unit of the differences is 10−4 , especially for MS-SSIMs shown in Table X.”
BackgroundMethodInfluentialBlind image quality assessment based on Benford's law
Al‐Bandawi, H., & Deng, G. (2018) · IET Image Processing · doi:10.1049/iet-ipr.2018.5385 · Cited by 6
Multivariate Statistical Approach to Image Quality Tasks
Gupta, P., Bampis, C. G., Glover, J. L., Paulter, N. G., & Bovik, A. C. (2018) · Journal of Imaging · doi:10.3390/jimaging4100117 · Cited by 6
Self-citationStereoscopic image quality assessment combining statistical features and binocular theory
Yang, J., Xu, H., Zhao, Y. S., Liu, H., & Lu, W. B. (2018) · Pattern Recognition Letters · doi:10.1016/j.patrec.2018.10.012 · Cited by 7
Aldahdooh, A., Masala, E., Wallendael, G. V., et al. (2019) · Signal Processing Image Communication · doi:10.1016/j.image.2019.01.004 · Cited by 6
Li, S., Peng, M., Zhang, B., et al. (2019) · International Journal of Remote Sensing · doi:10.1080/01431161.2019.1580793 · Cited by 6
HDR IMAGE QUALITY ASSESSMENT USING MACHINE-LEARNING BASED COMBINATION OF QUALITY METRICS
Choudhury, A., & Daly, S. J. (2018) · IEEE Global Conference on Signal and Information Processing · doi:10.1109/globalsip.2018.8646579 · Cited by 7
“While applying such front-end non-linearities to the LDR metrics does improve their performance (PLCC increases from 0.8635 to 0.9323 for MS-SSIM), the HDR metrics are in general more advanced, such as having orientation channels.”
MethodInfluentialPerceptually Based Adaptive Motion Retargeting to Animate Real Objects by Light Projection
Fukiage, T., Kawabe, T., & Nishida, S. (2019) · IEEE Transactions on Visualization and Computer Graphics · doi:10.1109/tvcg.2019.2898738 · Cited by 6
Dong, W. C., Bie, H., Lu, L., & Li, Y. (2019) · IEEE Access · doi:10.1109/access.2019.2922011 · Cited by 6
“, PSNR and MultiScale Structural SIMilarity (MSSSIM) [61], four state-of-the-art FR-IQA metrics, i.”
BackgroundInfluentialSparse Representation for No-Reference Quality Assessment of Satellite Stereo Images
Xiong, Y.-M., Shao, F., Meng, X.-C., Zhou, B., & Ho, Y.-S. (2019) · IEEE Access · doi:10.1109/access.2019.2932015 · Cited by 7
“For more information, see http://creativecommons.org/licenses/by/4.0/ 106295 of the distorted images after comparing them, including Structural SIMilarity index (SSIM) [9] and Multi-Mcale SSIM index (MS-SSIM) [10], assume that human visual system (HVS) tends to perceive the local structures in an image when evaluating…”
MethodInfluentialEncoding Distortions for Multi-task Full-Reference Image Quality Assessment
Huang, C., Jiang, T., & Jiang, M. F. (2019) · IEEE International Conference on Multimedia and Expo (ICME) · doi:10.1109/icme.2019.00320 · Cited by 6
“For fair comparison between learning based models and non-learning based models, we evaluate the performance of the non-learning based models including PSNR, MS-SSIM, FSIMc, VSI only on the test set.”
MethodSpeckle Noise Diffusion in Knee Articular Cartilage Ultrasound Images
Shoaib, M. A., Hossain, M. B., Hum, Y. C., et al. (2019) · Current Medical Imaging Formerly Current Medical Imaging Reviews · doi:10.2174/1573405615666190903143330 · Cited by 6
Quality-guided video aesthetics assessment with social media context
Zhang, C., Liu, S., & Li, H. (2019) · Journal of Visual Communication and Image Representation · doi:10.1016/j.jvcir.2019.102643 · Cited by 6
Variational approach for privacy funnel optimization on continuous data
Nan, L., & Tao, D. (2019) · Journal of Parallel and Distributed Computing · doi:10.1016/j.jpdc.2019.09.010 · Cited by 6
Zeeshan, M., & Majid, M. (2019) · Entropy · doi:10.3390/e21100964 · Cited by 6
“The MS − SSIM index between original and distorted videos is computed as, where l M ( Orig , Dist ) denotes luminance comparison, while c j ( Orig , Dist ) and s j ( Orig , Dist ) represent contrast and structure comparisons at j -th scale of original and distorted videos.”
MethodInfluentialImage quality assessment based on adaptive multiple Skyline query
He, S., & Liu, Z. (2019) · Signal Processing Image Communication · doi:10.1016/j.image.2019.115676 · Cited by 8
Zhou, Z., Li, J., Xu, Y., & Quan, Y. (2020) · Neural Computing and Applications · doi:10.1007/s00521-019-04694-9 · Cited by 6
“Many FR-IQA metrics characterize the structural information of images, such as SSIM [46], MS-SSIM [45], IW-SSIM [43] and FSIM [53], using manually designed features.”
MethodRecognizing Compressed Videos: Challenges and Promises
Pourreza, R., Ghodrati, A., & Habibian, A. (2019) · doi:10.1109/iccvw.2019.00129 · Cited by 6
“Subjective studies confirm that LPIPS better reflects human image quality measures compared to classical metrics such as PSNR and MSSSIM [34].”
BackgroundSubjective Evaluation of User Quality of Experience for Omnidirectional Video Streaming
Mrvelj, Š., Matulin, M., & Martirosov, S. (2020) · PROMET - Traffic&Transportation · doi:10.7307/ptt.v32i3.3444 · Cited by 9
Improvement of Learning Stability of Generative Adversarial Network Using Variational Learning
Lee, J.-Y., & Choi, S.-I. (2020) · Applied Sciences · doi:10.3390/app10134528 · Cited by 5
“Conversely, in comparison with α-GAN, α-GAN yields slightly higher MS–SSIM values compared to VLGAN, yet VLGAN is two times faster than α-GAN in terms of speed.”
MethodInfluentialImage Quality Assessment based on Dual Domains Fusion
Lu, Y., Li, W., Ning, X., Dong, X., Zhang, Y., & Sun, L. (2020) · International Conference on High Performance Big Data and Intelligent Systems (HPBD&IS) · doi:10.1109/hpbdis49115.2020.9130589 · Cited by 6
Metadata
- Authors
- Wang, Z., Simoncelli, E. P., & Bovik, A. C.
- Affiliations
- Courant Institute of Mathematical Sciences (US)
- The University of Texas at Austin (US)
- Venue
- The Thrity-Seventh Asilomar Conference on Signals, Systems & Computers, 2003
- Publisher
- IEEE
- Published
- 2004-07-08
- Pages
- 1398-1402
- ISSN
- 1058-6393 (ISSN-L)
- OpenAlex
- W1580389772
- Cited by
- 6,253
- References
- 15
- Citations by year
- 2026: 344 · 2025: 675 · 2024: 729 · 2023: 668 · 2022: 662 · 2021: 641 · 2020: 563 · 2019: 438 · 2018: 305 · 2017: 267 · 2016: 206 · 2015: 173 · 2014: 166 · 2013: 112 · 2012: 89
- FWCI
- 3.98
- Type
- conference paper
- Language
- en
- Open access
- closed
- Locations
- Asilomar Conference on Signals, Systems and Computers (ACSSC) (publishedVersion)
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Image and Video Quality Assessment
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Image Retrieval and Classification Techniques
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Digital Media Forensic Detection
- Keywords
- structural similarity, image quality assessment, multiscale analysis, human visual system, perceptual image quality, image generation, parameter calibration, viewing conditions
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
A language model drew the claims from the paper on 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.