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 3,201–3,300 of 6,047
Expediting structure–property analyses using variational autoencoders with regression
Templeton, W. F., Miner, J. P., Ngo, A., Fitzwater, L., Reddy, T., Abranovic, B., Prichard, P., Lewandowski, J. J., & Narra, S. P. (2024) · Computational Materials Science · doi:10.1016/j.commatsci.2024.113056 · Cited by 5
Computational single fundus image restoration techniques: a review
Zhang, S., Webers, C. A. B., & Berendschot, T. T. J. M. (2024) · Frontiers in Ophthalmology · doi:10.3389/fopht.2024.1332197 · Cited by 12
“The peak signal-to-noise ratio (PSNR) and structured similarity measure (SSIM) are reference-based metrics that can evaluate the image quality if the ground truth is known (73, 103 – 105).”
BackgroundVerma, K., Srivastava, S., & Mishra, R. K. (2024) · Multimedia Tools and Applications · doi:10.1007/s11042-024-19624-8 · Cited by 5
GOP-Based Deep Preprocessing for Video Coding
Arai, D., Iwamura, S., Iguchi, K., & Ichigaya, A. (2024) · Picture Coding Symposium · doi:10.1109/pcs60826.2024.10566387 · Cited by 6
Improving Deep Learning-based Saliency Detection Using Channel Attention Module
Farsi, H., Ghermezi, D., Barati, A., & Mohamadzadeh, S. (2024) · International Journal of Engineering · doi:10.5829/ije.2024.37.11b.21 · Cited by 10
Synthetic Face Discrimination via Learned Image Compression
Iliopoulou, S., Tsinganos, P., Ampeliotis, D., & Skodras, A. N. (2024) · Algorithms · doi:10.3390/a17090375 · Cited by 7
“The distortion refers to the expected difference between the original and reconstructed image, evaluated using a norm or perceptual metric function like mean squared error (MSE) or multi-scale structural similarity (MS-SSIM) [34].”
MethodComplexity Evaluation of Aerial Infrared Countermeasure Scenes
Xie, F., Dong, M., Wang, X., Yang, D., & Yan, J. (2024) · IEEE Transactions on Aerospace and Electronic Systems · doi:10.1109/taes.2024.3450453 · Cited by 6
Cai, S., Fang, F., & Wang, Y. (2024) · Physics of Fluids · doi:10.1063/5.0228384 · Cited by 5
Similarity characterization of geotechnical engineering sites under the site-specific data scenario
Han, L., Jiang, M., Zhang, W., et al. (2024) · Bulletin of Engineering Geology and the Environment · doi:10.1007/s10064-024-03990-6 · Cited by 5
Abad, A., Poux, A., & Brunel, M. (2024) · Journal of Quantitative Spectroscopy and Radiative Transfer · doi:10.1016/j.jqsrt.2024.109315 · Cited by 6
Tang, C., Eisenmenger, L., Rivera-Rivera, L., Huo, E., Junn, J., Kuner, A., Oechtering, T., Peret, A., Starekova, J., & Johnson, K. M. (2024) · Journal of Magnetic Resonance Imaging · doi:10.1002/jmri.29672 · Cited by 8
“7,8 10 – 12 To address these limitations, structural similarity (SSIM) 13,14 and perceptual loss 10 have been applied to DL reconstruction.”
Synthesizing CTA Image Data for Type-B Aortic Dissection using Stable Diffusion Models
Abaid, A., Farooq, M. A., Hynes, N., Corcoran, P., & Ullah, I. (2024) · Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC) · doi:10.1109/embc53108.2024.10782969 · Cited by 7
“We validated the effectiveness of the proposed method by statistically analysing the quality and diversity of synthetic images for each class.”
Multi-information guided camouflaged object detection
Shi, C., Zhao, L., Wang, R., et al. (2025) · Image and Vision Computing · doi:10.1016/j.imavis.2025.105470 · Cited by 5
Zheng, H., Sun, Y., Xu, H., Zhang, L., Han, Y., Cui, S. R., & Li, Z. (2025) · IEEE Journal of Oceanic Engineering · doi:10.1109/joe.2025.3529132 · Cited by 6
“L MS − SSIM can be expressed as follows [45]: where M is the total number of the scales, p = { p j : j = 1 , . . . , N 2 } , and g = { g j : j = 1 , . . . , N 2 } . μ p and μ g are the mean of p and g , σ p and σ g are the standard deviations of p and g , and σ pg represents their covariance. β m…”
Discriminator guided visible-to-infrared image translation
Ma, D., Su, J., Xian, Y., & Li, S. (2025) · Complex & Intelligent Systems · doi:10.1007/s40747-025-01827-7 · Cited by 5
Ramírez-Jaime, A., Arce, G. R., Porras-Diaz, N., Єремеєв, О., Rubel, A., Лукин, В. В., Kopytek, M., Lech, P., Fastowicz, J., & Okarma, K. (2025) · Remote Sensing · doi:10.3390/rs17071215 · Cited by 12
Lu, Z., Sun, H., Lei, L., Xu, Y., Sun, Y., & Kuang, G. (2025) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2025.3565657 · Cited by 9
“MS-SSIM [68], which stands for multiscale structural similarity index measure, enables a quality assessment of restored RSIs that is Authorized licensed use limited to the terms of the applicable license agreement with IEEE.”
SplatMAP: Online Dense Monocular SLAM with 3D Gaussian Splatting
Hu, Y. X., Liu, R., Chen, M., Beerel, P. A., & Feng, A. (2025) · Proceedings of the ACM on Computer Graphics and Interactive Techniques · doi:10.1145/3728310 · Cited by 10
Image Characteristic-Guided Learning Method for Remote-Sensing Image Inpainting
Zhou, Y., Gao, X., Wu, X., Wang, F., Jing, W., & Hu, X. (2025) · Remote Sensing · doi:10.3390/rs17132132 · Cited by 6
Zhang, J., Qin, P., Zeng, J., et al. (2025) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2025.3583382 · Cited by 5
“…performance indicators [6], including entropy (EN) [53], standard deviation (SD) [54], mutual information (MI) [55], multiscale SSIM (MS-SSIM) [56], sum of correlation di ff erences (SCDs) [57], spatial frequency (SF) [58], and average gradient (AG) [59], are used to compare the performance of…”
Image quality assessment based on structure variance classification
Zhan, Y., & Zhang, R. (2015) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp.2015.7178253 · Cited by 5
“The IQA methods are PSNR, SSIM [2], MS-SSIM [3], IW-SSIM [4], VIF [5], MAD [6], and DLAI [7].”
MethodInfluentialImage compression techniques using Local Binary Pattern
Szöke, I.-A., Lungeanu, D., & Holban, Ş. (2015) · IEEE International Symposium on Applied Machine Intelligence and Informatics (SAMI) · doi:10.1109/sami.2015.7061863 · Cited by 5
A computational texture masking model for natural images based on adjacent visual channel inhibition
Liu, Y., & Allebach, J. P. (2014) · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · doi:10.1117/12.2038599 · Cited by 5
Performance Evaluation of Distortion Measures for Retinal Images
Nirmala, S. R., Dandapat, S., & Bora, P. K. (2011) · International Journal of Computer Applications · doi:10.5120/2225-2835 · Cited by 5
Anand, A., & de Veciana, G. (2016) · International Symposium on Modeling and Optimization in Mobile, Ad Hoc, and Wireless Networks (WiOpt) · doi:10.1109/wiopt.2016.7492916 · Cited by 5
A SVR based quality metric for depth quality assessment
Zhang, Y., Jin, X., & Dai, Q. (2016) · IEEE International Symposium on Circuits and Systems proceedings · doi:10.1109/iscas.2016.7539117 · Cited by 5
“In order to demonstrate the efficiency of the proposed metric, we compare its performance with four state-of-the-art image quality objective models: peak signal-to-noise ratio (PSNR) [14], structural similarity index (SSIM) [15], multiscale SSIM (MSSSIM) [16], and Visual Information Fidelity (VIF) [17].”
MethodQuality assessment of MPEG-4 AVC/H.264 and HEVC compressed video in a telemedicine context
Chaabouni, A., Lambert, J., Gaudeau, Y., et al. (2017) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2017.8296926 · Cited by 5
Enhanced block truncation coding image using digital multitone screen
Guo, J. M., & Sankarasrinivasan, S. (2017) · Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC) · doi:10.1109/apsipa.2017.8282118 · Cited by 5
“Multi Scale SSIM is considered as the image quality assessment index [16].”
MethodImage quality assessment improvement via local gray-scale fluctuation measurement
Yang, X., Sun, Q., & Wang, T. (2018) · Multimedia Tools and Applications · doi:10.1007/s11042-018-5740-z · Cited by 6
“In addition to SSIM, there are many HVS-oriented IQA methods, such as the multiscale SSIM (MSSIM) [26], which extends the SSIM into multiscale space.”
MethodTotu, T., Buga, R. M., Dumitru, A. V., Costache, M., Sladoje, N., & Stanciu, S. G. (2018) · IEEE Access · doi:10.1109/access.2018.2868127 · Cited by 5
“(PSNR) [43], Structural Similarity Index (SSIM) [44], Pixel Based Visual Information Fidelity (VIFP) [45], Universal Quality Index (UQI) [46], [47], Noise Quality Measure (NQM) [48], Weighted Signal-to-Noise Ratio (WSNR) [49], Signal-to-Noise Ratio (SNR) [50], Multi-Scale SSIM (MS-SSIM) [51], Visual Signal To Noise Ra…”
MethodInfluentialLossy DCT-based compression of remote sensing images with providing a desired visual quality
Krivenko, S., Abramov, S., Лукин, В. В., et al. (2019) · doi:10.1117/12.2532726 · Cited by 5
Characterizing Generalized Rate-Distortion Performance of Video Coding: An Eigen Analysis Approach
Duanmu, Z., Liu, W., Li, Z., Ma, K., & Wang, Z. (2020) · IEEE Transactions on Image Processing · doi:10.1109/tip.2020.2988437 · Cited by 6
“Third, its precedent models SSIM [31] and MS-SSIM [32] have been demonstrated to perform well in estimating GRD functions [10], and have been widely used in industry practice.”
MethodSelf-citationEfficient in-situ image and video compression through probabilistic image representation
Liu, R., Li, M., & Ma, L. (2023) · Signal Processing · doi:10.1016/j.sigpro.2023.109268 · Cited by 8
Edge Guided Low-Light Image Enhancement
Rana, D., Lal, K. J., & Parihar, A. S. (2021) · International Conference on Intelligent Computing and Control Systems (ICICCS) · doi:10.1109/iciccs51141.2021.9432150 · Cited by 10
“• Structural Loss: Finally, we employ the Multiscale-SSIM (MS-SSIM) [21] metric to penalize loss of structural information.”
MethodAn Autoencoder with a Memory Module for Video Anomaly Detection
Tang, W., Feng, Y., & Li, J. (2021) · Youth Academic Annual Conference of Chinese Association of Automation (YAC) · doi:10.1109/yac53711.2021.9486538 · Cited by 5
“1) Multi-scale structure similarity: According to a topdown hypothesis proposed by [13], a human visual system is highly suitable for extracting structural information from the scene.”
BackgroundMethodA New Image Quality Index and it‟s Application on MRI Image
Islam, M. T., & Islam, S. M. R. (2021) · International Journal of Image Graphics and Signal Processing · doi:10.5815/ijigsp.2021.04.02 · Cited by 7
ES-CycleGAN: An Improved CycleGAN for VI-to-IR Translation
Su, T., Pan, F., & Cao, J. (2021) · Chinese Control Conference (CCC) · doi:10.23919/ccc52363.2021.9550245 · Cited by 5
Decode after filtering: a network for camouflage object segmentation
Zhang, C., Li, X., & Li, X. (2021) · Soft Computing · doi:10.1007/s00500-021-06388-8 · Cited by 5
“2017), SSIM loss function (Wang et al. 2003), which can be expressed by the following formulas:”
BackgroundMethodBVRIQE: A Completely Blind No Reference Virtual Reality Image Quality Evaluator
Poreddy, A. K. R., & Appina, B. (2022) · 2022 IEEE International Conference on Signal Processing and Communications (SPCOM) · doi:10.1109/spcom55316.2022.9840855 · Cited by 5
A Segmentation Algorithm of Colonoscopy Images Based on Multi-Scale Feature Fusion
Yu, J. J., Li, Z., Xu, C., & Feng, B. (2022) · Electronics · doi:10.3390/electronics11162501 · Cited by 5
Dual-Branch Remote Sensing Spatiotemporal Fusion Network Based on Selection Kernel Mechanism
Li, W., Wu, F., & Cao, D. (2022) · Remote Sensing · doi:10.3390/rs14174282 · Cited by 5
Enhanced Multiscale Attention Network for Single Image Dehazing
Imai, W., & Ikehara, M. (2022) · IEEE Access · doi:10.1109/access.2022.3204026 · Cited by 5
Stereoscopic video quality measurement with fine-tuning 3D ResNets
Imani, H., Islam, M. B., Junayed, M. S., Aydın, T., & Arıca, N. (2022) · Multimedia Tools and Applications · doi:10.1007/s11042-022-13485-9 · Cited by 5
“In [64], well-known 2D IQA methods such as SSIM, MS-SSIM [55], and VIF [43] are used to combine disparity and spatial information of the stereoscopic image to estimate the stereoscopic image quality.”
MethodMotion Compensation-based Low-Complexity Decoder Side Depth Estimation for MPEG Immersive Video
Garus, P., Henry, F., Maugey, T., & Guillemot, C. M. (2022) · IEEE International Workshop on Multimedia Signal Processing (MMSP) · doi:10.1109/mmsp55362.2022.9949346 · Cited by 5
“3 For each value s, we compute the amount of motion compensated blocks as well as the resulting synthesis PSNR and MS-SSIM [13].”
MethodRecognition-Aware Deep Video Compression for Remote Surveillance
Beye, F., Itsumi, H., Vitthal, C., et al. (2022) · Proceedings - International Conference on Image Processing · doi:10.1109/icip46576.2022.9897530 · Cited by 5
“In contrast to most prior works where MSE or MS-SSIM [9] are used as distortion measure in order to reduce the discrepancy between an uncompressed frame and its reconstruction, our approach can save bitrate by preventing details not relevant for object detection from being encoded.”
ResultFrame-Type Sensitive RDO Control for Content-Adaptive Encoding
Vibhoothi, V., Pitié, F., & Kokaram, A. C. (2022) · Proceedings - International Conference on Image Processing · doi:10.1109/icip46576.2022.9897204 · Cited by 5
“That rate measurement is derived from the MS-SSIM RD characteristic generated using N QP measurements.”
MethodInfluentialLocal Aggregative Attack on SAR Image Classification Models
Du, M., 毕大平, Du, M., Wu, Z., & Xu, X. (2022) · IEEE Advanced Information Technology, Electronic and Automation Control Conference (IAEAC) · doi:10.1109/iaeac54830.2022.9929962 · Cited by 6
“3) The structural dissimilarity (DSSIM) [18] metric is added to form the cost function such that it balances the constraint relationship between attack effectiveness and stealthiness.”
BackgroundConvolutional Residual Blocks With Edge Guidance for Image Denoising
Holla, K. S., & Lee, B. (2022) · International Conference on Information and Communication Technology Convergence (ICTC) · doi:10.1109/ictc55196.2022.9952960 · Cited by 5
Dwork, N., & Larson, P. E. Z. (2022) · Journal of Electronic Imaging · doi:10.1117/1.jei.31.6.063043 · Cited by 5
Fused features for no reference image quality assessment
Chemmanam, A. J., Shahanaz, N., & Jose, B. A. (2022) · The Imaging Science Journal · doi:10.1080/13682199.2023.2167267 · Cited by 7
“Visual Information Fidelity (VIF) [7], Structural Similarity Index (SSIM) [8], and it ’ s variants – Fast (FSIM) [9], Multi-Scale (M-SSIM) [10] etc. relates better with human visual perception.”
BackgroundNITS-IQA Database: A New Image Quality Assessment Database
Ruikar, J. D., & Chaudhury, S. (2023) · Sensors · doi:10.3390/s23042279 · Cited by 9
“Scatter plots of the subjective scores (MOS) vs. MSSIM.”
MethodChest X-Ray Quality Assessment Method With Medical Domain Knowledge Fusion
Du, S., Wang, Y., Huang, X., Zhao, R.-W., Zhang, X., Feng, R., Shen, Q., & Zhang, J. Q. (2023) · IEEE Access · doi:10.1109/access.2023.3252893 · Cited by 5
“There are already many FR-IQA methods, such as SSIM [5], MS-SSIM [6], VIF [7], IFC [8], FSIM [9], etc. Due to the existence of reference images, FR-IQA can usually accurately measure the quality of distorted images.”
MethodDeep polyp image enhancement using region of interest with paired supervision
Huang, D., Liu, J., Shi, Y., Li, C., & Tang, W. J. (2023) · Computers in Biology and Medicine · doi:10.1016/j.compbiomed.2023.106961 · Cited by 5
Super‐resolution of sodium images from simultaneous 1H MRF/23Na MRI acquisition
Rodriguez, G. G., Yu, Z., Shaykevich, S., et al. (2023) · NMR in Biomedicine · doi:10.1002/nbm.4959 · Cited by 5
“The M-SSIM can be more robust when compared with the SSIM concerning variations in viewing conditions.(23) Both SSIM and M-SSIM values would converge to 1 for an optimal generated image that is identical to the original or ground truth image.”
Automated large volume sample preparation for vEM
Stempinski, E., Pagano, L., Riesterer, J. L., Adamou, S., Thibault, G., Song, X. B., Chang, Y. H., & López, C. S. (2023) · Methods in cell biology · doi:10.1016/bs.mcb.2023.01.009 · Cited by 5
Varga, D. (2023) · Journal of Imaging · doi:10.3390/jimaging9060116 · Cited by 7
“The authors of multi-scale SSIM [8] extended the idea of SSIM into multiple scales.”
MethodInfluentialGu, J., Qiu, Q., Zhu, J., Cao, Q., Zhen, H., Li, B., & Shu, H. (2023) · Medical Physics · doi:10.1002/mp.16566 · Cited by 6
Shi, Y., Jiang, C., Liu, C., et al. (2023) · Electronics · doi:10.3390/electronics12132995 · Cited by 5
Single-pixel image reconstruction using coherent nonlinear optics
THOMAS, M., Kumar, S., & Huang, Y. (2023) · Optics Letters · doi:10.1364/ol.498296 · Cited by 5
Marie, A., Desnos, K., Morin, L., & Zhang, L. (2023) · International Conference on Quality of Multimedia Experience (QoMEX) · doi:10.1109/qomex58391.2023.10178503 · Cited by 5
“These metrics include similarity based metrics such as the SSIM [29] index and its multi-scale variant MS-SSIM [28], Feature SIM-ilarity (FSIM) [34], Spectral Residual based SIMilarity (SR-SIM) [32].”
MethodInfluentialLuna, R., Zabaleta, I., & Bertalmío, M. (2023) · Frontiers in Neuroscience · doi:10.3389/fnins.2023.1222815 · Cited by 8
“The results in Table 3 demonstrate that INRF-VQA consistently outperforms all state-of-the-art algorithms for all frame rates tested, including the highest frame rate of 120 fps (for a fair comparison, results for VMAF for 60 fps and 30 fps were also computed using frame dropping, and we do not rule out that other met…”
MethodScreen content video quality assessment based on spatiotemporal sparse feature
Ding, R., Zeng, H., Wen, H., Huang, H., Cheng, S., & Hou, J. (2023) · Journal of Visual Communication and Image Representation · doi:10.1016/j.jvcir.2023.103912 · Cited by 6
Disguised Heterogeneous Face Generation With Iterative-Adversarial Style Unification
Peng, C., Kong, Z., Liu, D., et al. (2023) · IEEE Transactions on Multimedia · doi:10.1109/tmm.2023.3314977 · Cited by 5
“We also use a number of image quality assessment metrics to evaluate the quality of our generated disguised faces, including the SSIM [54], MSSSIM [55], PSNR [56], LPIPS [57], FID [58], and FQA [59], which can help reflect the superiority of our method over the existing methods.”
MethodActive CT Reconstruction with a Learned Sampling Policy
Wang, C., Shang, K., Zhang, H., Zhao, S., Liang, D., & Zhou, S. K. (2023) · ACM International Conference on Multimedia (ACM MM) · doi:10.1145/3581783.3611746 · Cited by 8
“Reconstructed CT images are quantitatively measured by the multi-scale SSIM [33], [34], PSNR, and RMSE (Root Mean Square Error), and we compute the RMSE on normalized images.”
MethodHuang, W., Zhang, H., Cheng, Y., & Quan, X. (2023) · Frontiers in Physiology · doi:10.3389/fphys.2023.1241370 · Cited by 6
LFSphereNet: Real Time Spherical Light Field Reconstruction from a Single Omnidirectional Image
Gond, M., Zerman, E., Knorr, S. B., & Sjöström, M. (2023) · Conference on Visual Media Production · doi:10.1145/3626495.3626500 · Cited by 10
RITFusion: Reinforced Interactive Transformer Network for Infrared and Visible Image Fusion
Li, X., Li, Y., Chen, H., et al. (2023) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2023.3342223 · Cited by 5
“To objectively compare the fusion performance of different methods, the widely used metrics are employed for quantitative evaluation, including FMI [50], Q P [51], Q CV [52], and MS_SSIM [53].”
MethodXue, F., Guo, L., Bialkowski, A., & Abbosh, A. M. (2023) · Sensors · doi:10.3390/s24010008 · Cited by 6
“the neural network trained between 0.8 and 0.95, which can be seen as a good quality score [47]. are between 0.8 and 0.95, which can be seen as a good quality score [47] The proportions of SSIM values exceeding 0.9 in all the test datasets are relatively comparable, standing at 34.8%, 31.6%, and…”
BackgroundMethodInfluentialLi, F., Абрамов, С. К., Dohtiev, I., & Лукин, В. В. (2024) · Advanced Information Systems · doi:10.20998/2522-9052.2024.1.07 · Cited by 6
“The already performed studies show that there are “nonlinear” quality metrics [25] such as SSIM-based ones including MS-SSIM [26] and FSIM [27].”
BackgroundKapsner, L. A., Folle, L., Hadler, D., Eberle, J. S., Balbach, E. L., Liebert, A., Ganslandt, T., Wenkel, E., Ohlmeyer, S., Uder, M., et al. (2024) · Scientific Reports · doi:10.1038/s41598-024-56853-1 · Cited by 5
Attention-driven residual-dense network for no-reference image quality assessment
Zhang, Y., Song, Y., Lv, X., Shao, M., & Leng, Q. (2024) · Signal Image and Video Processing · doi:10.1007/s11760-024-03172-7 · Cited by 6
“The representative FR-IQA methods include structural similarity (SSIM) [1], multi-scale structural similarity (MS-SSIM) [2], information-weighted structural similarity (IW-SSIM) [3], feature similarity (FSIM) [4], visual saliency index (VSI) [5], gradient magnitude similarity deviation (GMSD) [6]…”
Indiscriminate Data Poisoning Attacks on Pre-trained Feature Extractors
Lu, Y., Yang, M. Y. R., Kamath, G., & Yu, Y. (2024) · IEEE Conference on Secure and Trustworthy Machine Learning (SaTML) · doi:10.1109/satml59370.2024.00023 · Cited by 13
“The model weights are obtained by training on SimCLR (Chen et al. 2020a) for 800 epochs with a batch size of 512.”
MethodHardware Implementation of a Deep Learning-based Model for Image Quality Assessment
Said, Y., & Alsariera, Y. A. (2024) · Engineering Technology & Applied Science Research · doi:10.48084/etasr.7194 · Cited by 6
Fast High Dynamic Range Radiance Fields for Dynamic Scenes
Wu, G., Yi, T., Fang, J., et al. (2024) · International Conference on 3D Vision (3DV) · doi:10.1109/3dv62453.2024.00052 · Cited by 5
“To assess the quality of the synthesized images from new viewpoints, we employ PSNR, SSIM [51], and LPIPS [58] metrics as evaluation measures.”
MethodA FUNQUE Approach to the Quality Assessment of Compressed HDR Videos
Venkataramanan, A. K., Stejerean, C., Katsavounidis, I. X., et al. (2024) · Picture Coding Symposium · doi:10.1109/pcs60826.2024.10566413 · Cited by 6
“By using the PU21 encoding function, several legacy video quality models such as SSIM [9], MS-SSIM [10], FSIM [11], and VSI [12] have been shown to be useful for HDR video quality assessment problems [13].”
MethodInfluentialSelf-citationBit Rate Matching Algorithm Optimization in JPEG-AI Verification Model
Jia, P., Koyuncu, A. B., Mao, J., et al. (2024) · Picture Coding Symposium · doi:10.1109/pcs60826.2024.10566454 · Cited by 5
“The loss is determined by comparing the reconstructed image with the original image and is a combination of the Mean Squared Error (MSE) [27] and the Multi-Scale Structural SIM-ilarity (MS-SSIM) [28] loss with varying weights.”
MethodYu, W., Nie, Y., & Xiong, W. (2024) · Asia Conference on Computer Vision, Image Processing and Pattern Recognition (CVIPPR) · doi:10.1145/3663976.3664001 · Cited by 5
Perceptual Crack Detection for Rendered 3D Textured Meshes
Sarvestani, A. S., Zhou, W. M., & Wang, Z. (2024) · International Conference on Quality of Multimedia Experience (QoMEX) · doi:10.1109/qomex61742.2024.10598253 · Cited by 8
“…integration method over six well-known base QA models, which include: (1) lumaPSNR, which computes the peak signal-to-noise ratio (PSNR) on the Y channels of reference and distorted frames in the YUV color space; (2) SSIM [19]; (3) MS-SSIM [27]; (4) IW-SSIM [20]; (5) FSIM [28]; (6) and FSIM C [28].”
MethodSelf-citationGAN-based Image Inpainting Techniques: A Survey
Mohod, A., & Nair, P. P. (2024) · International Conference on Advancements in Power, Communication and Intelligent Systems (APCI) · doi:10.1109/apci61480.2024.10616694 · Cited by 6
“• SSIM (Structural Similarity Index) [55] (2) where x , y and z are weighting factors that adjust the relative significance of the luminance, contrast, and structural comparisons in the overall SSIM calculation.”
BackgroundMethodLong-term reprojection loss for self-supervised monocular depth estimation in endoscopic surgery
Shi, X., Cui, B., Clarkson, M. J., & Islam, M. (2024) · Artificial Intelligence Surgery · doi:10.20517/ais.2024.17 · Cited by 5
A Study on the Effect of Color Spaces in Learned Image Compression
Prativadibhayankaram, S. V., Panda, M. P., Seiler, J., Richter, T., Sparenberg, H., Föessel, S., & Kaup, A. (2024) · Proceedings - International Conference on Image Processing · doi:10.1109/icip51287.2024.10648167 · Cited by 7
Kamışlı, F. (2024) · Multimedia Tools and Applications · doi:10.1007/s11042-024-20315-7 · Cited by 6
“) is any di ff erentiable distortion metric such as mean-squared-error (MSE) or multi-scale structural similarity index measure (MS-SSIM) [16]. g a U”
InfluentialLi, H., Xia, G., Yu, C., et al. (2024) · Sensors · doi:10.3390/s24216772 · Cited by 5
“Consequently, this paper selects Mean Squared Error (MSE) as the training metric, while evaluation of test outcomes is conducted using Peak Signal-to-Noise Ratio (PSNR) and Multi-Scale Structural Similarity (MS-SSIM) metrics [28].”
MethodA Variational Neural Network Based on Algorithm Unfolding for Image Blind Deblurring
Gong, S., Wang, Y., Yang, G., Wei, W., Zhao, J., & Pan, Z. (2024) · Applied Sciences · doi:10.3390/app142411742 · Cited by 5
STanH: Parametric Quantization for Variable Rate Learned Image Compression
Presta, A., Tartaglione, E., Fiandrotti, A., & Grangetto, M. (2025) · IEEE Transactions on Image Processing · doi:10.1109/tip.2025.3527883 · Cited by 15
“The image quality (i.e., distortion) is evaluated as peak signal-to-noise ratio (PSNR) and secondarily as multiscale structural similarity (MS-SSIM) [52].”
Catenary Clutter Elimination Network for Railway Tunnel Ground Penetrating Radar Data
Xiong, H., Li, J., Liu, T., et al. (2025) · IEEE Transactions on Geoscience and Remote Sensing · doi:10.1109/tgrs.2025.3533609 · Cited by 5
A deep learning-based approach for stegomalware sanitization in digital images
Liguori, A., Zuppelli, M., Gallo, D., et al. (2025) · Journal of Intelligent Information Systems · doi:10.1007/s10844-025-00936-6 · Cited by 5
“The SSIM is defined as: where μ x , μ ˆ x , σ 2 x , and σ 2 ˆ x are the mean intensities and the variances of images x and ˆ x , the σ x ˆ x is the covariance between the images, and C 1 and C 2 are small constants to stabilize the division (see, Wang et al. (2003) for further details).”
MSAENet: a multi-scale attention edge network for defect detection in strip steel surface defects
Deng, H., & Peng, J. (2025) · Measurement Science and Technology · doi:10.1088/1361-6501/adf500 · Cited by 5
“The hybrid loss function can be formally expressed as: where N = 6 represents the total number of side output layers. ℓ k cel , ℓ k bec , ℓ k iou , and ℓ k ssim represent CEL loss [29], BCE loss [46], IOU loss [47], and SSIM loss [48], respectively.”
Sun, W., Cao, L., Fu, K., Zhu, D., Jia, J., Hu, M., Min, X., & Zhai, G. (2025) · IEEE International Conference on Multimedia and Expo workshops · doi:10.1109/icmew68306.2025.11152231 · Cited by 10
“For example, Video BLIINDS [5] predicts perceptual video quality by extracting spatio-temporal natural scene statistics from frame-difference DCT coefficients and characterizing motion coherence and global motion.”
3DGS-IEval-15K: A Large-scale Image Quality Evaluation Database for 3D Gaussian-Splatting
Xing, Y., Wang, J., Niu, P., Huang, W., Zhai, G., & Xu, Y. (2025) · ACM International Conference on Multimedia (ACM MM) · doi:10.1145/3746027.3758206 · Cited by 10
Objective Quality Comparison of 4K UHD and Up-Scaled 4K UHD Videos
Cheon, M., & Lee, J. (2014) · IEEE International Symposium on Multimedia (ISM) · doi:10.1109/ism.2014.33 · Cited by 5
“We performed quality measurement using three fullreference quality metrics, namely, MS-SSIM [6], VIF [7], and VSNR [8].”
MethodInfluentialExponentiated combination of two scores for image quality assessment
Isono, K., & Shimamura, T. (2013) · IEEE Region 10 Conference (TENCON) · doi:10.1109/tencon.2013.6719032 · Cited by 5
Evaluating the multi-scale iCID metric
Moan, S. L., Preiss, J., & Urban, P. (2015) · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · doi:10.1117/12.2078879 · Cited by 5
Using full-reference image quality metrics for automatic image sharpening
Krasula, L., Fliegel, K., Callet, P. L., et al. (2014) · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · doi:10.1117/12.2052275 · Cited by 5
Image super-resolution via Kernel regression of sparse coefficients
Yuan, T., Zhou, F., Yang, W., et al. (2014) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp.2014.6854714 · Cited by 5
“And our method surpasses all the other methods in terms of PSNR, MS-SSIM and VIF.”
MethodUsing full reference image quality metrics to detect game engine artefacts
Piórkowski, R., & Mantiuk, R. (2015) · doi:10.1145/2804408.2804414 · Cited by 6
“2004], MSSIM (Multiscale Structural SIMilarity Index) [Wang et al. 2003], and HDR-VDP-2 (High Dynamic Range Visual Difference Predicator) [Mantiuk et al.”
BackgroundMethodInfluentialImage quality assessment using the amplitude/phase quantization code
Kim, D.-O., & Park, R. (2010) · IEEE Transactions on Consumer Electronics · doi:10.1109/tce.2010.5681166 · Cited by 6
“A multiscale SSIM (MSSSIM) was proposed [6].”
Filtering of dual-polarization radar images based on discrete cosine transform
Kozhemiakin, R. A., Лукин, В. В., Vozel, B., et al. (2014) · International Radar Symposium · doi:10.1109/irs.2014.6869301 · Cited by 5
Contourlet transform-based structural similarity for image quality assessment
Yang, C., Wang, F., & Xiao, D.-Q. (2009) · IEEE International Conference on Intelligent Computing and Intelligent Systems (ICIS) · doi:10.1109/icicisys.2009.5357699 · Cited by 5
Color face-tuned salient detection for image quality assessment
Tong, Y., Konik, H., & Trémeau, A. (2010) · European Workshop on Visual Information Processing · doi:10.1109/euvip.2010.5699149 · Cited by 5
“Many papers proposed image quality metrics, such as UQI, SSIM, LINLAB, PSNRHVS, PSNRHVSM, WSNR, IFC and VSNR, which give better results than PSNR and MSE [1-14].”
Result
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.