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 1,901–2,000 of 6,047
Perceptual Evaluation on Audio-Visual Dataset of 360 Content
Fela, R. F., Pastor, A., Callet, P. L., Zacharov, N., Vigier, T., & Forchhammer, S. O. (2022) · IEEE International Conference on Multimedia and Expo workshops · doi:10.1109/icmew56448.2022.9859426 · Cited by 29
“Lastly, we calculated 2D image quality metrics: PSNR, SSIM[38], and MS-SSIM[39].”
MethodSmart selective encryption of CAVLC for H.264/AVC video
Dubois, L., Puech, W., & Blanc-Talon, J. (2011) · IEEE International Workshop on Information Forensics and Security · doi:10.1109/wifs.2011.6123130 · Cited by 16
A SURVEY ON FULL REFERENCE IMAGE QUALITY ASSESSMENT ALGORITHMS
George, A., & Livingston, S. (2013) · International Journal of Research in Engineering and Technology · doi:10.15623/ijret.2013.0212052 · Cited by 20
“The multi scale structural similarity measure [7] i s a more flexible method than the other single scale methods .”
BackgroundMethodImage Denoising in the Transformed Domain Using Non Local Neighborhoods
Mseddi, W. S., Beghdadi, A., & Abed‐Meraim, K. (2006) · doi:10.1109/icassp.2006.1660481 · Cited by 16
A Structural Variation Classification Model for Image Quality Assessment
Zhan, Y.-B., Zhang, R., & Wu, Q. (2017) · IEEE transactions on multimedia · doi:10.1109/tmm.2017.2689923 · Cited by 15
CCTV quality assessment for forensics facial recognition analysis
Senan, M. F. E. M., Abdullah, S. N. H. S., Kharudin, W. M., & Saupi, N. A. M. (2017) · International Conference on Cloud Computing, Data Science & Engineering (Confluence) · doi:10.1109/confluence.2017.7943232 · Cited by 21
“Generally, video and image quality assessment algorithms are grouped into three categories: fullreference (FR), reduced-reference (RR), and noreference (NR) algorithms [5].”
BackgroundComparative Study of Dermoscopic Hair Removal Methods
Talavera-Martínez, L., Bibiloni, P., & González-Hidalgo, M. (2019) · Lecture notes in computational vision and biomechanics · doi:10.1007/978-3-030-32040-9_2 · Cited by 16
NAViDAd: A No-Reference Audio-Visual Quality Metric Based on a Deep Autoencoder
Martinez, H. B., Farias, M. C. Q., & Hines, A. (2019) · European Signal Processing Conference (EUSIPCO) · doi:10.23919/eusipco.2019.8902975 · Cited by 16
“0728 SSIM [12] PCC 0.”
PGGAN-Based Anomaly Classification on Chest X-Ray Using Weighted Multi-Scale Similarity
Kwon, H.-J., Shin, D.-H., & Chung, K. (2021) · IEEE Access · doi:10.1109/access.2021.3102954 · Cited by 17
A review on Video Quality Assessment
Nidhi, N., & Aggarwal, N. (2014) · doi:10.1109/raecs.2014.6799645 · Cited by 16
Checkerboard Context Model for Efficient Learned Image CompressionPreprint
He, D., Zheng, Y., Sun, B., et al. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2103.15306 · Cited by 16
“This work is done when Dailan He, Yaoyan Zheng and Baocheng Sun are interns at SenseTime Research. performance than those currently SOTA conventional compression techniques such as JPEG2000 [30] and BPG [8] on both the peak signal-to-noise ratio (PSNR) and multi-scale structural similarity (MS-SSIM) [36] distortion me…”
MethodInfluentialLou, A., Guan, S., & Loew, M. H. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2105.04075 · Cited by 16
“There are many popular methods to compare two images: • Binary to binary: Jaccard Similarity (JS) [37] • Gray to gray: Mean Absolute Error (MAE) [38]; Tanimoto Similarity [39] (extended Jaccard Similarity); Structural Similarity (SSIM) [40].”
MethodD4: Text-guided diffusion model-based domain adaptive data augmentation for vineyard shoot detection
Hirahara, K., Nakane, C., Ebisawa, H., Kuroda, T., Iwaki, Y., Utsumi, T., Nomura, Y., Koike, M., & Mineno, H. (2024) · Computers and Electronics in Agriculture · doi:10.1016/j.compag.2024.109849 · Cited by 20
PQMET: A digital image quality metric based on human visual system
Dimauro, G., Altomare, N., & Scalera, M. (2014) · International Conference on Image Processing Theory, Tools and Applications (IPTA) · doi:10.1109/ipta.2014.7002002 · Cited by 16
Multi‐style Chinese art painting generation of flowers
Fu, F., Lv, J., Tang, C., & Li, M. (2020) · IET Image Processing · doi:10.1049/ipr2.12059 · Cited by 17
Securing audio transmission based on encoding and steganography
Abood, E. W., Abduljabbar, Z. A., Sibahee, M. A. A., et al. (2021) · Indonesian Journal of Electrical Engineering and Computer Science · doi:10.11591/ijeecs.v22.i3.pp1777-1786 · Cited by 15
High Frequency Detail Accentuation in CNN Image Restoration
Ayyoubzadeh, S. M., & Wu, X. (2021) · IEEE Transactions on Image Processing · doi:10.1109/tip.2021.3120678 · Cited by 26
Hewage, C. T. E. R., Ahmad, A., Mallikarachchi, T., Barman, N., & Martini, M. G. (2022) · IEEE Access · doi:10.1109/access.2022.3180491 · Cited by 22
“The proposed scheme is proved to be better than the other two for all four IQM that were evaluated (PSNR, SSIM, multi-scale SSIM [80] and PSNR-HVS-M [81]).”
MethodAstrocytes Enhance Image Representation Encoded in Spiking Neural Network
Викторович, С. С., & Kazantsev, V. (2022) · Studies in computational intelligence · doi:10.1007/978-3-031-19032-2_20 · Cited by 15
Feng, X., Wang, T., Yang, X., Zhang, M., Guo, W., & Wang, W. (2022) · Mathematical Biosciences & Engineering · doi:10.3934/mbe.2023007 · Cited by 19
Synthetic PET via Domain Translation of 3-D MRI
Rajagopal, A., Natsuaki, Y., Wangerin, K. A., Hamdi, M., An, H., Sunderland, J. J., Laforest, R. A., Kinahan, P. E., Larson, P. E. Z., & Hope, T. A. (2022) · IEEE Transactions on Radiation and Plasma Medical Sciences · doi:10.1109/trpms.2022.3223275 · Cited by 20
“where μx and σ 2 x represent the mean and variance of volume x, μy and σ 2 y represent the mean and variance of volume y, σxy represents the covariance of x and y, and c∗ is chosen proportional to the dynamic range of pixel values [33].”
BackgroundLearned Disentangled Latent Representations for Scalable Image Coding for Humans and Machines
Özyılkan, E., Ulhaq, M., Choi, H., et al. (2023) · DCC · doi:10.1109/dcc55655.2023.00012 · Cited by 15
“Restrictions apply. structural similarity index (MS-SSIM) [7], which are used as distortion metrics between the original and reconstructed images within the loss function.”
MethodInfluentialPerceptual video quality assessment: the journey continues!
Saha, A., Pentapati, S. K., Shang, Z., Pahwa, R., Chen, B., Gedik, H. E., Mishra, S., & Bovik, A. C. (2023) · Frontiers in Signal Processing · doi:10.3389/frsip.2023.1193523 · Cited by 21
“Another notable resource on Cloud Gaming is the Tencent Gaming Video (TGV) dataset presented in Wen et al. (2021), however, the database is not publicly available.”
BackgroundSelf-citationZhi, S., Wang, Y., Xiao, H., Bai, T., Li, B., Tang, Y., Liu, C., Li, W., Li, T., Ge, H., et al. (2023) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2023.3294245 · Cited by 16
“Restrictions apply. where E indicates the expected value, and the MS-SSIM index is computed using the same parameters proposed in a previous study [47].”
MethodU-SMR: U-SwinT & multi-residual network for fabric defect detection
Qu, H., Di, L., Liang, J., & Liu, H. (2023) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2023.107094 · Cited by 19
Semisupervised Deep Neural Network-Based Cross-Frequency Ground-Penetrating Radar Data Inversion
Liu, H., Wang, J., Zhang, J., Jiang, H., Xu, J., Jiang, P., Zhang, F., Sui, Q., & Wang, Z. (2023) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2023.3316153 · Cited by 16
“A variety of evaluation indicators were used to quantitatively evaluate the effectiveness of our method, including MAE, mean squared error (mse), SSIM, and multiscale SSIM (MSSIM) [64].”
MethodSelf-citationTransformers and their application to medical image processing: A review
Zhu, D., & Wang, D. (2023) · Journal of Radiation Research and Applied Sciences · doi:10.1016/j.jrras.2023.100680 · Cited by 25
Liang, G., Zhou, J., Chen, Z., Wan, L., Wumener, X., Zhang, Y., Liang, D., Liang, Y., & Hu, Z. (2023) · EJNMMI Physics · doi:10.1186/s40658-023-00579-y · Cited by 20
“…measure (SSIM), the peak SNR (PSNR), the normalized mutual information (NMI), and their improved versions such as the multiscale SSIM (MS-SSIM) [30], information content-weighted SSIM (IW-SSIM) [31], feature similarity index measure (FSIM) [32], spectral residual-based similarity index measure…”
BackgroundCai, Z., Ma, Y., Huang, J., Mei, X., & Fan, F. (2023) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2023.3341137 · Cited by 22
“…fusion performance through eight metrics, where are average gradient (AG) [47], entropy (EN) [48], standard deviation (SD) [49], edge information ( Q AB / F ) [50], spatial frequency (SF) [51], visual information fidelity (VIF) [52], structural similarity (SSIM) [53], and multiscale SSIM (Qp) [54].”
BackgroundOne Transform to Compute Them All: Efficient Fusion-Based Full-Reference Video Quality Assessment
Venkataramanan, A. K., Stejerean, C., Katsavounidis, I. X., & Bovik, A. C. (2023) · IEEE Transactions on Image Processing · doi:10.1109/tip.2023.3345227 · Cited by 24
“A key weakness of SOTA quality models such as MS-SSIM and VMAF is that they operate only on the luma channel.”
MethodInfluentialSelf-citationExploring Long- and Short-Range Temporal Information for Learned Video Compression
Wang, H., & Chen, Z. (2024) · IEEE Transactions on Image Processing · doi:10.1109/tip.2024.3349859 · Cited by 18
“Besides, it is hard to optimize the hybrid coding framework towards complex objective metrics such as MS-SSIM [4] and LPIPS [5].”
MethodRan, G., Yao, X., Wang, K., Ye, J., & Ou, S. (2024) · Measurement Science and Technology · doi:10.1088/1361-6501/ad1eb6 · Cited by 18
“…difference heatmaps for the images generated by our proposed method, we also used three numerical metrics for quantitative model evaluation, including freshet inception distance (FID) [60], multi-scale structural similarity (MS-SSIM) [61], and learned perceptual image patch similarity (LPIPS) [62].”
Temporal context video compression with flow-guided feature prediction
Wang, Y., Huang, Q., Tang, B., Sun, H., Guo, X., & Miao, Z. (2024) · Expert Systems with Applications · doi:10.1016/j.eswa.2024.123322 · Cited by 18
Xu, T., Da, Y., Yang, G., et al. (2024) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2024.3359351 · Cited by 15
“We have also introduced a loss function that combines mean absolute error (MAE) and mean structural similarity index (MS-SSIM) [24], [25] losses, which can be expressed as Using this loss function as an additional feature extractor, we calculate the perceptual loss to assist in distinguishing real…”
MethodSemantic Communications with Explicit Semantic Base for Image Transmission
Zheng, Y., Wang, F., Xu, W., Pan, M., & Zhang, P. (2023) · Globecom · doi:10.1109/globecom54140.2023.10437876 · Cited by 17
“…schemes under a given CBR, which by first the channel capacity C is obtained as and then SNR is obtained according to the Shannon channel capacity formula, Peak signal-to-noise ratio (PSNR) and multi-scale structural similarity (MS-SSIM) [26] are used to measure the quality of image reconstruction.”
MethodAdaptive Multi-Source Predictor for Zero-Shot Video Object Segmentation
Zhao, X., Chang, S., Pang, Y., Yang, J., Zhang, L., & Lu, H. (2024) · International Journal of Computer Vision · doi:10.1007/s11263-024-02024-8 · Cited by 19
“For the depth stream, we follow the related works [42, 53, 57] to adopt the combination loss of L1 and SSIM [75].”
Lv, S., Wen, J., & Zhang, X. (2024) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2024.3428644 · Cited by 15
“We used the multiscale version of SSIM (MS-SSIM) [43] to offer a more detailed assessment of the degree of local structural likeness.”
MethodLocalized Plasmonic Structured Illumination Microscopy Using Hybrid Inverse Design
Wu, Q., Xu, Y., Zhao, J., Liu, Y., & Liu, Z. (2024) · Nano Letters · doi:10.1021/acs.nanolett.4c03069 · Cited by 15
Zhang, B., Wang, X., Ding, L., Zang, Q., Cong, B., Cai, H., Liu, T., & Zhong, Y. (2024) · Construction and Building Materials · doi:10.1016/j.conbuildmat.2024.139081 · Cited by 15
Su, H., Liu, Y., Liu, Q., Yuan, H., & Hamzaoui, R. (2025) · IEEE Transactions on Visualization and Computer Graphics · doi:10.1109/tvcg.2025.3532651 · Cited by 31
“Then, existing image quality assessment methods, such as SSIM [29], MS-SSIM [30], IW − SSIM p [31], VIFP [32], and UNIQUE [33] are used to predict the quality of the point cloud from the average quality of the projection images.”
Assessing the QoE in Video Services Over Lossy Networks
Suárez, F. J., Mangas, A. G., Granda, J. C., Garcia, D. F., & Nuño, P. (2015) · Journal of Network and Systems Management · doi:10.1007/s10922-015-9343-y · Cited by 16
Quality Assessment of Images with Multiple Distortions using Combined Metrics
Okarma, K. (2014) · Elektronika ir Elektrotechnika · doi:10.5755/j01.eee.20.6.7284 · Cited by 16
Stereoscopic 3D image quality assessment based on cyclopean view and depth map
Fezza, S. A., & Larabi, M. (2014) · IEEE International Conference on Consumer Electronics - Berlin (ICCE-Berlin) · doi:10.1109/icce-berlin.2014.7034289 · Cited by 15
“Even tough several metrics can be used, we opted for the Multi-Scale Structural SIMilarity (MS-SSIM) metric [24] for its simplicity and availability:”
MethodBIQWS: efficient Wakeby modeling of natural scene statistics for blind image quality assessment
Jenadeleh, M., & Moghaddam, M. E. (2016) · Multimedia Tools and Applications · doi:10.1007/s11042-016-3785-4 · Cited by 16
“Examples of such algorithms are structural similarity index (SSIM) [49], multiscale SSIM (MS-SSIM) [48], visual information fidelity (VIF) [40], feature similarity index (FSIM) [55] and recent Visual Saliency based Index method named VSI [56].”
BackgroundMethodInfluentialQuality assessment of 3D synthesized images via disoccluded region discovery
Zhou, Y., Li, L.-D., Gu, K., Fang, Y., & Lin, W. (2016) · International Conference on Information Photonics · doi:10.1109/icip.2016.7532510 · Cited by 13
“The 2D image quality metrics include PSNR, SSIM [12], MS-SSIM [13], IW-SSIM [14], VSNR [15], MAD [16], FSIM [17], VSI [18], GSM [19], GMSD [20], ADD-SSIM [21], while four DIBR metrics are VSQA [2], 3Dswim [3], MW-PSNR [4], MP-PSNR [5].”
MethodQuadratic trigonometric B-spline for image interpolation using GA
Hussain, M. Z., Abbas, S., & Irshad, M. (2017) · PLoS ONE · doi:10.1371/journal.pone.0179721 · Cited by 15
Zhu, M., & He, B. (2017) · Optical Engineering · doi:10.1117/1.oe.56.11.113105 · Cited by 15
García, M., Segura-García, J., & Felici-Castell, S. (2017) · Computer Communications · doi:10.1016/j.comcom.2017.12.010 · Cited by 15
No-Reference Stereoimage Quality Assessment for Multimedia Analysis Towards Internet-of-Things
Yang, J., Jiang, B., Song, H., Yang, X., Lu, W. B., & Liu, H. (2018) · IEEE Access · doi:10.1109/access.2018.2791560 · Cited by 15
“It has been found that images are naturally multiscale and the scale we choose can affect the performance of the IQA algorithm [32], [33].”
BackgroundCenter-Emphasized Visual Saliency and a Contrast-Based Full Reference Image Quality Index
Layek, M., Uddin, A. S., Le, T. P., Chung, T., & Huh, E. (2018) · Symmetry · doi:10.3390/sym11030296 · Cited by 13
“nsider HVS, poor correlation with humans, low computation, widely used. [3] SSIM Luminance, contrast, structure Average Milestone method to consider structural information that better represents HVS. [4] MS-SSIM Multi-scale structure Weighted sum An extension to SSIM that incorporates variations of viewing conditions…”
MethodAdaptive compression method for underwater images based on perceived quality estimation
Ya-Qiong, C., Zou, H., & Yuan, F. (2019) · Frontiers of Information Technology & Electronic Engineering · doi:10.1631/fitee.1700737 · Cited by 15
Feng, X., Su, X., Shen, J., & Jin, H. (2019) · Remote Sensing · doi:10.3390/rs11161910 · Cited by 15
“It is defined as the ratio between the maximum possible value of a signal and the value of distorting noise that affects the quality of its representation [36].”
BackgroundSimple Yet Effective Way for Improving the Performance of Lossy Image Compression
Yeo, Y.-J., Shin, Y.-G., Sagong, M.-C., Kim, S., & Ko, S.-J. (2020) · IEEE Signal Processing Letters · doi:10.1109/lsp.2020.2982561 · Cited by 16
“…(BSD) [17], and the LIVE dataset [18], the performance of the proposed method is compared with that of an existing method in terms of image quality assessment (IQA) metrics: peak signal-to-noise ratio (PSNR), structural similarity (SSIM) [19], and multi-scale structural similarity (MS-SSIM) [20].”
MethodCloud-based elastic architecture for distributed video encoding: Evaluating H.265, VP9, and AV1
Aguado, J., Peña-Ortiz, R., García, M., et al. (2020) · Journal of Network and Computer Applications · doi:10.1016/j.jnca.2020.102782 · Cited by 15
Discriminative Sketch Topic Model With Structural Constraint for SAR Image Classification
Zhang, Y., Liu, F., Jiao, L., Yang, S., Li, L., & Yang, M. (2020) · IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing · doi:10.1109/jstars.2020.3024002 · Cited by 15
“We use SSIM [51] to measure the structural similarity of sketch maps x i and x j , that is, the structuralsimilaritiesofimages d i and d j . σ 1 isthemeansquared distance of all structural similarities SSIM ( x i , x j ) , and σ 2 is the mean squared distance between two texture features y i and y…”
MethodSubjective And Objective Quality Assessment Of Mobile Gaming Video
Wen, S., Ling, S., Wang, J., Chen, X., Jing, Y., & Callet, P. L. (2022) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp43922.2022.9746547 · Cited by 15
“Apart from those NR metrics, 4 commonly used, especially in treaming industry [2], full reference metrics were also tested, including the Peak Signal to Noise Ratio (PSNR) [26], the Structural Similarity Index (SSIM) [27], the Multi-Scale Structural Similarity (MS-SSIM) [28], and the Video Multi-Method Assessment Fusi…”
MethodComputational interference microscopy enabled by deep learning
Jiao, Y., He, Y. R., Kandel, M. E., Liu, X., Lu, W., & Popescu, G. (2021) · APL Photonics · doi:10.1063/5.0041901 · Cited by 15
Liu, L., Sang, X., Yu, X., et al. (2021) · Optics Communications · doi:10.1016/j.optcom.2021.127032 · Cited by 15
Li, L., Tang, J., Shao, Z., Tan, X., & Ma, L. (2021) · The Visual Computer · doi:10.1007/s00371-021-02188-1 · Cited by 16
“To quantitatively evaluate our method, we utilize three popular metrics: Frechet Inception Distance (FID) [14], Inception Score (IS) [33] and Multi-Scale Structural Similarity (MS-SSIM) [41].”
MethodCompression efficiency analysis of AV1, VVC, and HEVC for random access applications
Nguyen, T. L., & Marpe, D. (2021) · APSIPA Transactions on Signal and Information Processing · doi:10.1017/atsip.2021.10 · Cited by 27
“This imbalance for XPSNR, VMAF, and MS-SSIM is significant for the sequences ParkRunning3 and BQTerrace, where the latter shows a bit-rate saving for AV1 relative to HM when using VMAF, whereas a bit-rate overhead for AV1 relative to HM has been measured when using MS-SSIM.”
BackgroundInfluentialBASNet: A Boundary-Aware Siamese Network for Accurate Remote-Sensing Change Detection
Wei, H., Chen, R., Yu, C., et al. (2021) · IEEE Geoscience and Remote Sensing Letters · doi:10.1109/lgrs.2021.3119885 · Cited by 15
DeSVQ: Deep Learning Based Streaming Video QoE Estimation
Ghosh, M., Singhal, C., & Wayal, R. (2022) · International Conference on Distributed Computing and Networking (ICDCN) · doi:10.1145/3491003.3491023 · Cited by 20
“We have used selective numerical features, i.e., Peak-Signal-toNoise-Ratio (PSNR), Multi-Scale Structural Similarity Index MSSSIM [29], Spatio-Temporal Reduced Reference Entropic Differencing (STRRED) [27], and Video Multi-method Assessment Fusion (VMAF) [23] as input to the LSTM network (in the second stage) that is…”
MethodRubel, A., Єремеєв, О., Лукин, В. В., Fastowicz, J., & Okarma, K. (2022) · Applied Sciences · doi:10.3390/app12041986 · Cited by 21
“, Structural Similarity—SSIM [14,15]) metrics can be applied.”
BackgroundContent-Oriented Learned Image Compression
Li, M., Gao, S., Feng, Y., et al. (2022) · Lecture notes in computer science · doi:10.1007/978-3-031-19800-7_37 · Cited by 15
Video authentication in digital forensic
Gupta, A., Gupta, S., & Mehra, A. (2015) · International Conference on Futuristic Trends on Computational Analysis and Knowledge Management · doi:10.1109/ablaze.2015.7154945 · Cited by 17
A GPU-Based Algorithm-Specific Optimization for High-Performance Background Subtraction
Zhang, C., Tabkhi, H., & Schirner, G. (2014) · Proceedings of the International Conference on Parallel Processing · doi:10.1109/icpp.2014.27 · Cited by 17
“Throughout all optimizations, the output quality remains largely untouched with -4% in foreground quality (MS-SSIM[11]) compared to singlethreaded execution.”
MethodEdge-based objective evaluation of image quality
Bondžulić, B. P., & Petrovic, V. S. (2011) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2011.6116378 · Cited by 15
Islamian, J. P., Toossi, M. T. B., Momennezhad, M., et al. (2012) · World Journal of Nuclear Medicine · doi:10.4103/1450-1147.103419 · Cited by 15
Evaluation of image quality using dual-tree complex wavelet transform and compressive sensing
Kim, D.-O., & Park, R. (2010) · Electronics Letters · doi:10.1049/el.2010.3452 · Cited by 15
A new full-reference image quality metric based on just noticeable difference
Toprak, S., & Yalman, Y. (2017) · Comput. Stand. Interfaces · doi:10.1016/j.csi.2016.08.003 · Cited by 12
An FPGA Implementation of a Convolutional Auto-Encoder
Zhao, W., Jia, Z., Wei, X., & Wang, H. (2018) · Applied Sciences · doi:10.3390/app8040504 · Cited by 16
Robust destriping method based on data-driven learning
Kuang, X., Sui, X., Liu, Y., Liu, C., Chen, Q., & Gu, G. (2018) · Infrared Physics & Technology · doi:10.1016/j.infrared.2018.09.015 · Cited by 16
Speeding Up VP9 Intra Encoder With Hierarchical Deep Learning-Based Partition Prediction
Paul, S., Norkin, A., & Bovik, A. C. (2020) · IEEE Transactions on Image Processing · doi:10.1109/tip.2020.3011270 · Cited by 17
“By instead optimizing the partition decision process in terms of a suitable perceptual quality model like SSIM [45], MS-SSIM [46], VIF [47], ST-RRED [48], or VMAF [49], [50], RD performance could potentially be further improved, which is a direction that we intend to explore as part of our future…”
MethodSelf-citationConvolutional Neural Network Training for RGBN Camera Color Restoration Using Generated Image Pairs
Han, Z., Li, L., Jin, W., et al. (2020) · IEEE photonics journal · doi:10.1109/jphot.2020.3025088 · Cited by 15
“Its equation is [29] ⎧⎪⎪⎪⎪⎪⎪⎪⎨ ⎪⎪⎪⎪⎪⎪⎪⎩ SSIM(f, g) = l (f, g)c(f, g)s(f, g) l (f, g) = 2μfμg +C1 μf + μg +C1 c(f, g) = 2σfσg +C2 σ 2 f + σ 2 g +C2 s(f, g) = σf g +C3 σfσg +C3 , (12)”
Sun, W., Tang, C., Li, W., Yuan, Z., Yang, H., & Liu, Y. (2020) · Lecture notes in computer science · doi:10.1007/978-3-030-58577-8_15 · Cited by 17
“Our video codec outperforms the standard H.264, H.265 and DL-based method [26] in compression quality measured by peak signal-to-noise ratio (PSNR) and multiscale structural similarity (MS-SSIM) [24].”
MethodCapturing Video Frame Rate Variations via Entropic Differencing
Madhusudana, P. C., Birkbeck, N., Wang, Y., Adsumilli, B., & Bovik, A. C. (2020) · IEEE Signal Processing Letters · doi:10.1109/lsp.2020.3028687 · Cited by 16
“…we employ a local Mean Subtracted (MS) filtering similar to [21] be the reference and distorted MS coefficients where local mean is calculated as [11] 0.4898 0.3407 0.4673 10.726 FSIM [13] 0.4469 0.3151 0.4435 10.874 ST-RRED [20] 0.5531 0.3800 0.5107 10.431 SpEED [21] 0.4861 0.3409 0.4449 10.866…”
InfluentialSelf-citationBlind Image Deblurring via Local Maximum Difference Prior
Liu, J., Tan, J., He, L., Ge, X., & Hu, D. (2020) · IEEE Access · doi:10.1109/access.2020.3039281 · Cited by 15
“We compute the kernel similarity [47] and the energy value (15) at the finest image scale.”
MethodA Method for Reduced-Reference Color Image Quality Assessment
Yu, M., Liu, H., Guo, Y., & Zhao, D. (2009) · International Congress on Image and Signal Processing (CISP) · doi:10.1109/cisp.2009.5300959 · Cited by 16
How to assess image quality within a workflow chain: an overview
Ciocca, G., Corchs, S., Gasparini, F., & Schettini, R. (2014) · International Journal on Digital Libraries · doi:10.1007/s00799-014-0124-0 · Cited by 18
“[118] An extension of the SSIM.”
A Novel Image Quality Index for Image Quality Assessment
Islam, S. M. R., Huang, X., & Le, K. (2013) · Lecture notes in computer science · doi:10.1007/978-3-642-42051-1_68 · Cited by 15
“[5] and [6] measure structure based on a spatially localized measure of correlation in pixel values structural similarity (SSIM) and in wavelet coefficients MS-SSIM.”
BackgroundMethodHybrid regularizers-based adaptive anisotropic diffusion for image denoising
Liu, K., Tan, J., & Ai, L. (2016) · SpringerPlus · doi:10.1186/s40064-016-1999-6 · Cited by 15
Learning Digital Camera Pipeline for Extreme Low-Light ImagingPreprint
Zamir, S. W., Arora, A., Khan, S., et al. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1904.05939 · Cited by 15
“In this work, we utilize the multi-scale structural similarity measure (MS-SSIM) [42]: LMS-SSIM(ŷ,y) = 1− 1 N N∑ p=1 MS-SSIM(ŷp,yp).”
MethodInfluentialA Survey of Safety and Trustworthiness of Deep Neural NetworksPreprint
Huang, X., Kroening, D., Ruan, W., et al. (2018) · arXiv (Cornell University) · Cited by 15
SENSEI: Aligning Video Streaming Quality with Dynamic User SensitivityPreprint
Zhang, X., Ou, Y., Sen, S., & Jiang, J. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2008.04687 · Cited by 28
“een the encoded frames and the original (uneconded) frames. Metrics of pixel-based quality, sometimes called visual quality assessment (VQA), include quantization parameter (QP) [45], PSNR [30], SSIM [56,57], STRRED [49], VMAF [11] and DeepVQA [35]. Streaming quality incidents include events during the streaming proce…”
BackgroundChen, E., Guo, M., Deng, J., Tian, Y., Deng, X. F., Le, J., & Zhang, H. (2024) · Expert Systems with Applications · doi:10.1016/j.eswa.2024.124290 · Cited by 15
Enhancing fairness in AI-enabled medical systems with the attribute neutral framework
Hu, L., Li, D., Liu, H., Chen, X., Gao, Y., Huang, S., Peng, X., Zhang, X., Bai, X., Yang, H., et al. (2024) · Nature Communications · doi:10.1038/s41467-024-52930-1 · Cited by 25
On the use of deep learning for computer-generated holography
Yu, X., Zhang, H., Zhao, Z., Fan, X., Hu, S., Li, Z., Chen, W., Li, D., Shi, S., Xiong, W., et al. (2025) · iScience · doi:10.1016/j.isci.2025.112507 · Cited by 21
“Afterward, some variant indexes are proposed: MS-SSIM extends SSIM by enhancing multiscale analysis of texture and structure, 70 and FSIM evaluates feature similarity by comparing the intensity distribution of image features at corresponding locations.”
HyPSAM: Hybrid Prompt-Driven Segment Anything Model for RGB-Thermal Salient Object Detection
Hou, R., Li, X., Ren, T., Zhou, D., Wu, G., & Cao, J. (2025) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2025.3613770 · Cited by 33
“Additionally, the other decoders are supervised by the hybrid loss [53]: where ℓ bce , ℓ ssim and ℓ iou denote BCE loss [56], SSIM loss [57] and IoU loss [58], respectively.”
Remote Sensing Image Enhancement Based on Non-Local Means Filter in NSCT Domain
Li, L., Si, Y., & Jia, Z. (2017) · Algorithms · doi:10.3390/a10040116 · Cited by 14
A Perceptually Inspired New Blind Image Denoising Method Using $L_{1}$ and Perceptual Loss
Uddin, A. F. M. S., Chung, T., & Bae, S. (2019) · IEEE Access · doi:10.1109/access.2019.2926848 · Cited by 14
“[37] suggests that, instead of using L2, a combination of L1 and Multi-Scale SSIM (MS-SSIM) [31], which is a variation of SSIM, can improve the performance of image restoration methods.”
MethodAssessing a Single Image in Reference-Guided Image Synthesis
Guo, J., Du, C., Wang, J., Huang, H., Wan, P., & Huang, G. M. (2022) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v36i1.19956 · Cited by 20
“Widely-used FR-IQA metrics include MS-SSIM (Wang, Simoncelli, and Bovik 2003), SSIM (Wang et al. 2004), PSNR (Huynh-Thu and Ghanbari 2008), FSIM (Zhang et al. 2011) and LPIPS (Zhang et al. 2018a).”
Image Quality Assessment with Gradient Siamese Network
Cong, H., Fu, L., Zhang, R., Zhang, Y., Wang, H., He, J.-R., & Gao, J. (2022) · 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw56347.2022.00127 · Cited by 8
“[45] 0.666 0.301 / 0.364 0.759 0.364 / 0.395 UQI [46] 0.965 0.460 / 0.504 0 .869 0.419 / 0.450 SSIM [14] 0.696 0.319 / 0.377 0.752 0.361 / 0.391 MS-SSIM [24] 0.457 0.337 / 0.119 0.532 0.369 / 0.163 RFSIM [47] 0.539 0.253 / 0.285 0.632 0.303 / 0.328 GSM [48] 0.828 0.378 / 0.450 0 .874 0.409 / 0.464 SRSIM [49] 1.155 0.5…”
MethodInfluentialResting state network mapping in individuals using deep learning
Luckett, P. H., Lee, J. J., Park, K. Y., Raut, R. V., Meeker, K. L., Gordon, E. M., Snyder, A. Z., Ances, B. M., Leuthardt, E. C., & Shimony, J. S. (2023) · Frontiers in Neurology · doi:10.3389/fneur.2022.1055437 · Cited by 19
“MSSI estimates similarity of a pair of 3D volumes by first estimating similarity of the pair for successively down-sampled versions of the pair, typically down-sampling by factors of 2, 4, 8, 16, and 32.”
MethodInfluentialFast fetal head compounding from multi-view 3D ultrasound
Wright, R., Gómez, A., Zimmer, V. A., Toussaint, N., Khanal, B., Matthew, J., Skelton, E., Kainz, B., Rueckert, D., Hajnal, J. V., et al. (2023) · Medical Image Analysis · doi:10.1016/j.media.2023.102793 · Cited by 24
Iterative Retinex-Based Decomposition Framework for Low Light Visibility Restoration
Gasparyan, H. A., Hovhannisyan, S. A., Babayan, S. V., & Agaian, S. С. (2023) · IEEE Access · doi:10.1109/access.2023.3269719 · Cited by 15
“We adopted seven image quality metrics (both reference and no reference) to evaluate the performance of RSD-Net [49], [56], [57].”
MethodLow-Light Image Enhancement Using a Simple Network Structure
Matsui, T., & Ikehara, M. (2023) · IEEE Access · doi:10.1109/access.2023.3290490 · Cited by 17
“We introduce a combination of L1 loss and Multi-Scale Structural Similarity (MS-SSIM) loss [33].”
MethodZeng, L., Yang, N., Li, X., Chen, A., Jing, H., & Zhang, J. (2023) · Electronics · doi:10.3390/electronics12183808 · Cited by 15
“In contrast to Baluja’s MSE algorithm [22], which only considers pixel-wise errors, Zhang et al. introduced the Structural Similarity Index Metric (SSIM) [24] and Multi-Scale Structure Similarity Index Metric (MSSIM) [25] into the loss function to account for the basic image structure.”
BackgroundMethodInfluentialUnderwater Image Quality Assessment from Synthetic to Real-world: Dataset and Objective Method
Li, X., Xu, H., Jiang, G., Yu, M., Luo, T., Zhang, X., & Ying, H. (2023) · ACM Transactions on Multimedia Computing Communications and Applications · doi:10.1145/3624983 · Cited by 19
“Current research on IQA has shown that adding multiscale information while evaluating performance results in IQA that correlates better with human perception [47, 48].”
BackgroundDeep learning in computed tomography super resolution using multi‐modality data training
Fok, W. Y. R., Fieselmann, A., Herbst, M., Ritschl, L., Kappler, S. G., & Saalfeld, S. (2023) · Medical Physics · doi:10.1002/mp.16825 · Cited by 20
Face De-Morphing Based on Diffusion Autoencoders
Long, M., Yao, Q., Zhang, L., & Peng, F. (2024) · IEEE Transactions on Information Forensics and Security · doi:10.1109/tifs.2024.3359029 · Cited by 21
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.