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 2,401–2,500 of 6,047
Wan, W., Huang, D., Shang, B., Wei, S., Wu, H. R., Wu, J., & Shi, G. (2022) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2022.3165970 · Cited by 11
“Due to the lake of 3D depth quality assessment metrics and correlation between the depth quality and the video quality, several classical 2D FR metrics, including PSNR, SSIM, MSSSIM [45] and FSIM [46], are used to evaluate the depth quality of S3D video, which are selected to compare the performances with the proposed…”
Method3-D Acoustic Image Denoising for a Sonar System With Sparse Planar Arrays
Zhao, D., Cai, T., Chen, P., Hu, Y., Liang, S., Mao, W., Liang, R., & Guo, X. (2022) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2022.3184345 · Cited by 11
Efficient and Adaptable Patch-Based Crack Detection
Guo, J.-M., & Markoni, H. (2022) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2022.3181475 · Cited by 11
“The last loss is the SSIM [40] which is denoted as L SSIM as in Eq.”
Guo, Q., & Peng, H. (2022) · Evolving Systems · doi:10.1007/s12530-022-09460-2 · Cited by 11
“The SSIM index can be defined as (Oliva et al. 2014): where I is the average of I , I ′ is the average of the division with weak denominator, L is the dynamic range of the pixel-values, k 1 = 0.01 and k 2 = 0.03 by default (Wang et al. 2003).”
BackgroundInfluentialSelf-supervised monocular Depth estimation with multi-scale structure similarity loss
Han, C., Cheng, D., Kou, Q., Wang, X., Chen, L., & Zhao, J. (2022) · Multimedia Tools and Applications · doi:10.1007/s11042-022-14012-6 · Cited by 10
“Inspired by the literature [33], the weight W is defined in this paper, and W = [0.”
BackgroundMethodInfluentialCNN Model with Multilayer ASPP and Two-Step Cross-Stage for Semantic Segmentation
Park, M.-H., Cho, J.-H., & Kim, Y.-T. (2023) · Machines · doi:10.3390/machines11020126 · Cited by 10
Zhang, H., Su, S., Zhu, Y., Sun, J., & Zhang, Y. (2023) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp49357.2023.10095465 · Cited by 15
“The comparison falls into three categories: FR-IQA (PSNR, SSIM [2], MS-SSIM [20], FSIM [21], GS [22], and GMSD [23]), NR-IQA (NIQE [10], BRISQUE [24], BLIINDS-II [25], DIIVINE [26], PaQ-2-PiQ [27], and MEON [28]), and SR-IQA (SIS [3], NSS-SR [17], DeepSRQ [6], DISQ [7], and DISQ+).”
Visual Security Index Combining CNN and Filter for Perceptually Encrypted Light Field Images
Wen, W.-Y., Huang, M.-H., Zhang, Y., Fang, Y., & Zuo, Y.-F. (2023) · ACM Trans. Multim. Comput. Commun. Appl. · doi:10.1145/3612924 · Cited by 9
“D (i.e., PSNR, SSIM, FSIM, MS-SSIM, VIF, VSI, GMSD, LPIPS, and DISTS) and the models for LF images (i.e., Wen and LFG-LFC).”
MethodInfluentialDeep learning assisted microwave-plasma interaction based technique for plasma density estimation
Ghosh, P. K., Chaudhury, B., Purohit, S., et al. (2023) · Journal of Physics D Applied Physics · doi:10.1088/1361-6463/acfdb6 · Cited by 10
BVI-VFI: A Video Quality Database for Video Frame Interpolation
Danier, D., Zhang, F., & Bull, D. (2023) · IEEE Transactions on Image Processing · doi:10.1109/tip.2023.3327912 · Cited by 19
“Additionally, we evaluated other popular conventional image quality metrics including MS-SSIM [69], VIF [70], VSI [21], FSIM [110], GMSD [111], and deep learning-based ones: DISTS [22], PieAPP [112], DeepIQA [113] and CONTRIQUE [74].”
MethodNIF: A Fast Implicit Image Compression with Bottleneck Layers and Modulated Sinusoidal Activations
Catania, L., & Allegra, D. (2023) · ACM International Conference on Multimedia (ACM MM) · doi:10.1145/3581783.3613834 · Cited by 12
VRL-IQA: Visual Representation Learning for Image Quality Assessment
Aslam, M. A., Xu, W., Ahmed, N., Saleem, G., Amin, T., & Caixue, H. (2023) · IEEE Access · doi:10.1109/access.2023.3340266 · Cited by 13
“More advanced methods like the Structural Similarity Index Metric (SSIM) [2] and its derivatives such as Multi-Scale SSIM (MS-SSIM) [3], three-component SSIM (3)- SSIM) [4], Complex Wavelet SSIM (CW-SSIM) [5], Information Content Weighted SSIM (IW-SSIM) [6], along with other modern alternatives,…”
MethodHIDRO-VQA: High Dynamic Range Oracle for Video Quality Assessment
Saini, S., Saha, A., & Bovik, A. C. (2024) · IEEE/CVF Winter Conference on Applications of Computer Vision Workshops (WACVW) · doi:10.1109/wacvw60836.2024.00056 · Cited by 16
Self-citationHDR or SDR? A Subjective and Objective Study of Scaled and Compressed Videos
Ebenezer, J. P., Shang, Z., Chen, Y., Wu, Y., Wei, H., Sethuraman, S., & Bovik, A. C. (2024) · IEEE Transactions on Image Processing · doi:10.1109/tip.2024.3404890 · Cited by 15
“SSIM (MS-SSIM) [48], HDR VDP 3 [36] Video Multimethod Assessment Fusion (VMAF) [23], Spatio-Temporal Reduced Reference Entropic Differences (STRRED) [25], Spatial Efficient Entropic Differencing (SpEED) [24], and the Space-Time GeneRalized Entropic Difference (STGREED) [26] video quality models on the newly created da…”
MethodSelf-citationLossy Compression of Single-channel Noisy Images by Modern Coders
Kryvenko, S., Лукин, В. В., & Vozel, B. (2024) · Remote Sensing · doi:10.3390/rs16122093 · Cited by 10
PIRN: Phase Invariant Reconstruction Network for infrared image super-resolution
Dan, J., Jin, T., Chi, H., Liu, M., Yu, J., Cao, K., Xinjing, Y., Luo, Z., & Xie, H. (2024) · Neurocomputing · doi:10.1016/j.neucom.2024.128221 · Cited by 11
Distributed Deep Joint Source-Channel Coding with Decoder-Only Side Information
Yilmaz, S. F., Özyılkan, E., Gündüz, D., & Erkip, E. (2024) · IEEE International Conference on Machine Learning for Communication and Networking (ICMLCN) · doi:10.1109/icmlcn59089.2024.10625214 · Cited by 14
“Unlike the MSE distortion metric (which is applied point-wise), perception-oriented distortion criteria, such as LPIPS and multi-scale SSIM (MS-SSIM) [31], measure the similarity between the original image and the reconstructed one using a proxy for the human perception.”
MethodOn the application of multivariate kernel density estimation to image error concealment
Koloda, J., Peinado, A. M., & Sánchez, V. (2013) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp.2013.6637867 · Cited by 10
“In order to reflect the perceptual quality of the reconstructions, the multi-scale structural similarity (MS-SSIM) index [18] is used for comparison along with the objective PSNR measure.”
MethodBlind image quality assessment for noise
Liu, M., Zhai, G., Zhang, Z., Sun, Y., Gu, K., & Yang, X. (2014) · IEEE International Symposium on Broadband Multimedia Systems and Broadcasting (BMSB) · doi:10.1109/bmsb.2014.6873480 · Cited by 11
Yan, P., Mou, X., & Xue, W. (2015) · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · doi:10.1117/12.2083283 · Cited by 13
“Successful FR-IQA includes GMSD ([14]), SSIM ([15]), FSIM ([16]), rNSE ([17]), MS-SSIM ([18]), VSNR ([19]), MAD ([20]), and the VIF index ([21]).”
MethodInfluentialA subjective evaluation of true 3D images
Tamboli, R. R., Vupparaboina, K. K., Ready, J., Jana, S., & Channappayya, S. S. (2014) · International Conference on D Imaging · doi:10.1109/ic3d.2014.7032603 · Cited by 10
“First, we calculate pairwise objective scores, namely MS-SSIM [7], and PSNR between the original image and the corresponding distorted image for each distortion type.”
MethodInfluentialLee, H., Lee, K. H., Kim, K. J., et al. (2010) · Medical Physics · doi:10.1118/1.3457471 · Cited by 10
LoG acts as a good feature in the task of image quality assessment
Mou, X., Xue, W., Chen, C., & Zhang, L. (2014) · Electronic imaging · doi:10.1117/12.2038982 · Cited by 12
“We also compare them with PSNR and other state of art models, including IFC [13], VIF [14], MAD [5], NSER [8], MS-SSIM [7], SSIM [3], IW-SSIM [16], and FSIM [2].”
MethodQian, F., Guo, J. B., Sun, T., & Wang, T. (2014) · Optics & Laser Technology · doi:10.1016/j.optlastec.2014.10.001 · Cited by 10
Integrating Visual Saliency Information into Objective Quality Assessment of Tone-Mapped Images
Liu, X., Zhang, L., Li, H., et al. (2014) · Lecture notes in computer science · doi:10.1007/978-3-319-09333-8_41 · Cited by 10
Natural motion statistics for no-reference video quality assessment
Saad, M. A., & Bovik, A. C. (2009) · International Conference on Quality of Multimedia Experience (QoMEX) · doi:10.1109/qomex.2009.5246957 · Cited by 10
Self-citationGMM Estimation of 2D-RCA Models With Applications to Texture Image Classification
Boulemnadjel, A., Hachouf, F., & Kharfouchi, S. (2015) · IEEE Transactions on Image Processing · doi:10.1109/tip.2015.2494740 · Cited by 10
Lossy compression of Landsat multispectral images
Kozhemiakin, R. A., Abramov, S., Лукин, В. В., et al. (2016) · Mediterranean Conference on Embedded Computing (MECO) · doi:10.1109/meco.2016.7525714 · Cited by 10
Morphology based iterative back-projection for super-resolution reconstruction of image
Nayak, R., Harshavardhan, S., & Patra, D. (2014) · doi:10.1109/et2ecn.2014.7044978 · Cited by 10
“Comparison of PSNR and UQI using different SR reconstruction methods are represented in Table 1-2 respectively.”
Kuppusamy, P. G., Joseph, J., & Sivaraman, J. (2017) · Journal of Applied Biomedicine · doi:10.1016/j.bbe.2016.12.008 · Cited by 12
Video quality assessment: A review of full-referenced, reduced-referenced and no-referenced methods
Soong, H.-C., & Lau, P.-Y. (2017) · IEEE International Colloquium on Signal Processing & Its Applications (CSPA) · doi:10.1109/cspa.2017.8064957 · Cited by 10
Brunet, D., Channappayya, S. S., Wang, Z., Vrscay, E. R., & Bovik, A. C. (2017) · Springer eBooks · doi:10.1007/978-3-319-61609-4_2 · Cited by 12
“The Multi-Scale-SSIM (MS-SSIM) index [41] is an improvement over the SSIM index in that it measures structural similarity over multiple spatial scales.”
MethodSelf-citationA JND-Based Pixel-Domain Algorithm and Hardware Architecture for Perceptual Image Coding
Wang, Z., Tran, T.-H., Muthappa, P. K., et al. (2019) · Journal of Imaging · doi:10.3390/jimaging5050050 · Cited by 10
“It can be seen from the MS-SSIM and HDR-VDP curves that the perceptual codec exhibited a clear gain in perceptual quality over JPEG-LS in a rate range between 1 and 3.5 bits-per-pixel (bpp).”
MethodInfluentialSelf-citation3D Photoacoustic Tomography System Based on Full-View Illumination and Ultrasound Detection
Sun, M., Hu, D., Zhou, W., et al. (2019) · Applied Sciences · doi:10.3390/app9091904 · Cited by 10
Toward a Quality Predictor for Stereoscopic Images via Analysis of Human Binocular Visual Perception
Liu, Y., Kong, F., & Zhen, Z. (2019) · IEEE Access · doi:10.1109/access.2019.2919155 · Cited by 10
“There are many pooling strategies, such as PSNR, SSIM [54], MS-SSIM [55], and ADD-GSIM [56].”
MethodDeep Inverse Tone Mapping for Compressed Images
Wang, C., Zhao, Y., & Wang, R. (2019) · IEEE Access · doi:10.1109/access.2019.2920951 · Cited by 10
“We achieve the highest score of HDR-VDP-2, PU-PSNR, PU-MS_SSIM, log_PSNR while the lowest weber_MSE, which means the results from our method have high fidelity.”
MethodVOIDGA: A View-Approximation Oriented Image Database Generation Approach
Lukasczyk, J., Kinner, E. G., Ahrens, J. P., et al. (2018) · IEEE Symposium on Large Data Analysis and Visualization · doi:10.1109/ldav.2018.8739204 · Cited by 10
“The MS-SSIM, on the other hand, requires input parameters that can currently only be chosen heuristically [41].”
BackgroundMethodInfluentialCost-driven framework for progressive compression of textured meshes
Portaneri, C., Alliez, P., Hemmer, M., et al. (2019) · ACM Multimedia Systems Conference (MMSys) · doi:10.1145/3304109.3306225 · Cited by 10
“For this paper, we use the image-based perceptual metric detailed in [7] which compares image renderings of the textured mesh with MS-SSIM [22].”
Robust Cloth Warping via Multi-Scale Patch Adversarial Loss for Virtual Try-On Framework
Ayush, K., Jandial, S., Chopra, A., Hemani, M., & Krishnamurthy, B. (2019) · IEEE International Conference on Computer Vision Workshops · doi:10.1109/iccvw.2019.00161 · Cited by 11
“Table 1 summarizes the performance of our proposed approach against CP-VTON on benchmark metrics for image quality [1, 5, 4, 2] and pair-wise structural similarity [7, 8].”
MethodVideo compression to support the expansion of whole-slide imaging into cytology
Zarella, M. D., & Jakubowski, J. (2019) · Journal of medical imaging · doi:10.1117/1.jmi.6.4.047502 · Cited by 10
A Total Variation Denoising Method Based on Median Filter and Phase Consistency
Huang, S., & Wan, S. (2020) · Sensing and Imaging · doi:10.1007/s11220-020-00281-8 · Cited by 11
“To further compare these results, mean structural similarity index measurement (MSSIM) [22, 25-29] and SSIM index map [25, 29] are employed, as shown in Equations (17) and (18), where σimage is the standard variation of image, cov(uN(j), uclean(j)) denotes the covariance of images in square windows centered by pixel j…”
MethodHigh Resolution Zero-Shot Domain Adaptation of Synthetically Rendered Face Images
Garbin, S. J., Kowalski, M., Johnson, M., & Shotton, J. (2020) · Lecture notes in computer science · doi:10.1007/978-3-030-58604-1_14 · Cited by 11
“Having obtained a sequence of proposals, from step 1-4, we simply select the one that matches input most using the Structural Similarity (SSIM) [38] metric at a resoluton of 368 pixels, which we empirically found to give better qualitative results than the LPIPS distance.”
MethodA Novel Multi-Focus Image Fusion Network with U-Shape Structure
Pan, T., Jiang, J., Yao, J., et al. (2020) · Sensors · doi:10.3390/s20143901 · Cited by 10
“The SSIM loss of two images x and y is defined as: lSSIM = 1− (2µxµy + C1)(2σxy + C2) (µ2x + µ 2 y + C1)(σ2x + σ2y + C2) , (5) where µx and µy are the mean of x and y, σx and σx denote the standard deviations of x and y, and σxy is their covariance.”
MethodInfluentialPerceptually Inspired Weighted MSE Optimization Using Irregularity-Aware Graph Fourier Transform
Lu, K.-S., Ortega, A., Mukherjee, D., & Chen, Y. (2020) · Proceedings - International Conference on Image Processing · doi:10.1109/icip40778.2020.9190876 · Cited by 13
Comprehensive performance analysis of objective quality metrics for digital holography
Ahar, A., Birnbaum, T., Chlipała, M., et al. (2021) · Signal Processing Image Communication · doi:10.1016/j.image.2021.116361 · Cited by 10
“The tested IQMs include: FSIM[58], IWSSIM[59], MS-SSIM[60], VIF[61], NLP[62], GMSD[63].”
Consecutive Context Perceive Generative Adversarial Networks for Serial Sections Inpainting
Zhang, S., Wang, L., Zhang, J., Gu, L., Jiang, X., Zhai, X., Sha, X., & Chang, S. (2020) · IEEE Access · doi:10.1109/access.2020.3031973 · Cited by 10
“…the quantitative evaluation using 5 metrics on testing datasets: Structural similarity index (SSIM) [40], Multi-scale structural similarity (MS-SSIM) [41], Feature Similarity Index (FSIM) [42], Peak signal-to-noise ratio (PSNR) [43] and Learned perceptual image patch similarity (LPIPS) [44].”
BackgroundMethodTowards Better Quality Assessment of High-Quality Videos
Ling, S., Baveye, Y., Nandakumar, D., et al. (2020) · doi:10.1145/3423328.3423496 · Cited by 10
“ThePeakSignaltoNoiseRatio (PSNR)[30], Structural Similarity Index (SSIM) [31] Multi-Scale Structural Similarity (MS-SSIM) [32], and Video Multi-Method Assessment Fusion (VMAF) [16] are four commonly used VQA metrics in the streaming industry.”
MethodDerivation Procedure of Coefficients of Metadata-Based Model for Adaptive Bitrate Streaming Services
Yamagishi, K., Egi, N., YOSHIMURA, N., et al. (2021) · IEICE Transactions on Communications · doi:10.1587/transcom.2020cqp0002 · Cited by 10
“The VMAF algorithm is widely used as it covers a wide range of applications and has estimation accuracy [3], [28] that comparable to that of human perception and overwhelming higher than those of structural similarity index measure (SSIM) [29] and multiscale SSIM (MS-SSIM) [30].”
Method3D Noise and Adversarial Supervision Is All You Need for Multi-modal Semantic Image Synthesis
Sushko, V., Schönfeld, E., Zhang, D., et al. (2020) · Lecture notes in computer science · doi:10.1007/978-3-030-65414-6_39 · Cited by 10
“We evaluate quantitatively using the Fréchet Inception Distance (FID) [3], mean Intersection-over-Union (mIoU) and MS-SSIM [9] with LPIPS [11] for multi-modal synthesis.”
MethodInfluentialOn the performance of objective quality metrics for lightfields
Mahmoudpour, S., & Schelkens, P. (2021) · Signal processing. Image communication · doi:10.1016/j.image.2021.116179 · Cited by 10
Real-Time, Deep Synthetic Aperture Sonar (SAS) Autofocus
Gerg, I. D., & Monga, V. (2021) · IEEE International Geoscience and Remote Sensing Symposium (IGARSS) · doi:10.1109/igarss47720.2021.9554141 · Cited by 10
“The measures evaluated are peak signal-to-noise ratio (PSNR) [22] and multiscale structural similarity (MS-SSIM) [23].”
BackgroundMethodBlind 360-degree image quality assessment via saliency-guided convolution neural network
Qiu, M., & Shao, F. (2021) · Optik · doi:10.1016/j.ijleo.2021.166858 · Cited by 14
Toward Identity Preserving Face Synthesis Between Sketches and Photos Using Deep Feature Injection
Lin, Y., Fu, K., Ling, S., Wang, J., & Cheng, P. (2021) · IEEE Transactions on Industrial Informatics · doi:10.1109/tii.2021.3074989 · Cited by 10
Chetouani, A., & Pedersen, M. (2021) · Applied Sciences · doi:10.3390/app11104661 · Cited by 12
“In Table 4, the correlation coefficients for each distortion, at five levels, are shown and compared to those of MSSIM and CNN QUALITY.”
MethodInfluentialReducing magnetic resonance image spacing by learning without ground-truth
Xuan, K., Si, L., Zhang, L., Xue, Z., Jiao, Y., Yao, W., Shen, D. G., Wu, D., & Wang, Q. (2021) · Pattern Recognition · doi:10.1016/j.patcog.2021.108103 · Cited by 17
FOVQA: Blind Foveated Video Quality Assessment
Jin, Y., Patney, A., Webb, R., & Bovik, A. C. (2022) · IEEE Transactions on Image Processing · doi:10.1109/tip.2022.3185738 · Cited by 12
“FR models have been extensively studied and used, and are well exemplified by SSIM [8] and MS-SSIM [9], whereby perceptually relevant luminance, contrast, and structure comparison measurements are integrated.”
MethodSelf-citationVideo Quality Assessment With Serial Dependence Modeling
Liu, Y., Wu, J., Li, A., Li, L., Dong, W., Shi, G., & Lin, W. (2021) · IEEE Transactions on Multimedia · doi:10.1109/tmm.2021.3107148 · Cited by 12
A Novel on Conditional Min Pooling and Restructured Convolutional Neural Network
Park, J., Kim, J.-Y., Huh, J., Lee, H., Jung, S.-H., & Sim, C.-B. (2021) · Electronics · doi:10.3390/electronics10192407 · Cited by 12
Sun, W., & Chen, Z. (2022) · IEEE Transactions on Image Processing · doi:10.1109/tip.2022.3142999 · Cited by 10
Efficient Generation of Cancelable Face Templates Based on Quantum Image Hilbert Permutation
Alhumyani, H. A., El‐Banby, G. M., El‐sayed, H. S., et al. (2022) · Electronics · doi:10.3390/electronics11071040 · Cited by 10
“Mathematically, SSIM can be computed as in Equation (7) [62] to evaluate the distortion strength.”
MethodKölln, L. S., Salem, O., Valli, J., Hansen, C. G., & McConnell, G. (2022) · Journal of Cell Science · doi:10.1242/jcs.258994 · Cited by 10
Simulating the perceptual effects of electrode–retina distance in prosthetic vision
Avraham, D., & Yitzhaky, Y. (2022) · Journal of Neural Engineering · doi:10.1088/1741-2552/ac6f82 · Cited by 11
Time Dependent Image Generation of Plants from Incomplete Sequences with CNN-Transformer
Drees, L., Weber, I., Rußwurm, M., & Roscher, R. (2022) · Lecture notes in computer science · doi:10.1007/978-3-031-16788-1_30 · Cited by 11
Improvements to a temporal filter for video coding
Enhorn, J., Hollmann, C., Sjöberg, R., et al. (2022) · IEEE International Symposium on Circuits and Systems proceedings · doi:10.1109/iscas48785.2022.9937263 · Cited by 10
Expanded Adaptive Scaling Normalization for End to End Image Compression
Shin, C., Lee, H., Son, H., Lee, S., Lee, D., & Lee, S. (2022) · Lecture notes in computer science · doi:10.1007/978-3-031-19790-1_24 · Cited by 11
“180] for MSE distortion loss, and to [7, 15, 30, 48, 110, 220] for MS-SSIM distortion loss.”
BackgroundMethodSubjective Quality Assessment of V-PCC-Compressed Dynamic Point Clouds Degraded by Packet Losses
Dumić, E., & Cruz, L. A. D. S. (2023) · Sensors · doi:10.3390/s23125623 · Cited by 10
“• MSE (Mean Squared Error), implementation from Matlab 64-bit 2020a; • PSNR (Peak Signal to Noise Ratio), implementation from Matlab 64-bit 2020a; • PSNRHVS (Peak Signal to Noise Ratio—Human Visual System) [36], implementation from [37]; • PSNRHVSM (Peak Signal to Noise Ratio—Human Visual System—Modified) [38], implem…”
MethodBlind Omnidirectional Image Quality Assessment: Integrating Local Statistics and Global Semantics
Zhou, W. J., & Wang, Z. (2023) · Proceedings - International Conference on Image Processing · doi:10.1109/icip49359.2023.10222049 · Cited by 22
“To capture the multi-scale characteristics of the human visual system [10], we construct pyramid representations [11, 12] of multiple local viewports.”
MethodSelf-citationSubjective and objective quality assessment of videos in error-prone network environments
Kim, S., Chae, C., & Lee, J. (2015) · Multimedia Tools and Applications · doi:10.1007/s11042-015-2613-6 · Cited by 10
“– Multi-Scale Structural SIMilarity index (MS-SSIM) is an extended version of SSIM, where multiple levels of decomposition are considered [38].”
BackgroundYou, C., Liu, Y., Zhao, B., et al. (2014) · Journal of Software · doi:10.4304/jsw.9.4.1050-1054 · Cited by 10
Monogenic signal theory based feature similarity index for image quality assessment
Luo, X., Wang, H., & Wang, S. (2014) · AEU - International Journal of Electronics and Communications · doi:10.1016/j.aeue.2014.07.015 · Cited by 10
“Recently, numerous extensions of SSIM have been developed, like Multi-Scale SSIM (MS-SSIM) index[4] , complex wavelet structural similarity(CW-SSIM) index[5] and four-component SSIM[5], which to some extent improve the performance of measure.”
DSIM: A DisSIMilarity-Based Image Clutter Metric for Targeting Performance
Xu, D., & Shi, Z. (2013) · IEEE Transactions on Image Processing · doi:10.1109/tip.2013.2270112 · Cited by 12
Image quality assessment based on local edge direction histogram
Chen, X., Zhang, R., & Zheng, S. (2011) · International Conference on Image Analysis and Signal Processing · doi:10.1109/iasp.2011.6109008 · Cited by 10
“However, it is noticed that SSIM is less competitive when used to measure the qualities of blurred and white noisy images.”
BackgroundSegmentation-based Perceptual Image Quality Assessment (SPIQA)
Ghanem, B., Resendiz, E., & Ahuja, N. (2008) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2008.4711774 · Cited by 10
No-reference quality assessment of contrast-distorted images
Xu, M. G., & Wang, Z. (2016) · IEEE International Conference on Signal and Image Processing (ICSIP) · doi:10.1109/siprocess.2016.7888285 · Cited by 11
“FR IQA methods [1] [2] [3] [4] [5] are the most reliable because they get complete information of reference images.”
MethodSelf-citationZhou, Y., Yang, L., Li, L., Gu, K., & Tang, L. (2017) · Multimedia Tools and Applications · doi:10.1007/s11042-017-5543-7 · Cited by 10
“The perception of image details is influenced by several factors, such as the distance between the test image and the subjects and the sampling density of the image information [46].”
BackgroundMethodInfluentialYing, C., Zhao, P., & Li, Y. (2018) · Journal of Electronic Imaging · doi:10.1117/1.jei.27.1.013026 · Cited by 10
Modeling Continuous Video QoE Evolution: A State Space Approach
Eswara, N., Sethuram, H. P., Chakraborty, S., Kuchi, K., Kumar, A., & Channappayya, S. S. (2018) · IEEE International Conference on Multimedia and Expo (ICME) · doi:10.1109/icme.2018.8486557 · Cited by 9
“[16] provides a comprehensive study of various VQA metrics and suggests that the metrics MS-SSIM [12] and MOVIE [13] provide good video quality prediction performances.”
BackgroundMethodChahid, A., Serrai, H., Achten, E., et al. (2018) · doi:10.1109/embc.2018.8513615 · Cited by 10
“The image denoising performance has been evaluated using two standard metrics: the Structural SIMilarity index (SSIM) [9] which evaluates the preservation of image details, the Global Phase Coherence (GPC) to measure the sharpness of the image [16].”
MethodMeasuring CT Reconstruction Quality with Deep Convolutional Neural Networks
Patwari, M., Gutjahr, R., Raupach, R., et al. (2019) · Lecture notes in computer science · doi:10.1007/978-3-030-33843-5_11 · Cited by 11
On the accuracy of video quality measurement techniques
Nandakumar, D., Wu, Y., Wei, H., et al. (2019) · IEEE International Workshop on Multimedia Signal Processing (MMSP) · doi:10.1109/mmsp.2019.8901796 · Cited by 10
Adversarial Video Compression Guided by Soft Edge Detection
Kim, S.-S., Park, J., Bampis, C. G., Lee, J., Markey, M. K., Dimakis, A. G., & Bovik, A. C. (2020) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp40776.2020.9054165 · Cited by 13
Self-citationLearning to compress videos without computing motion
Chen, M., Goodall, T., Patney, A., & Bovik, A. C. (2022) · Signal Processing Image Communication · doi:10.1016/j.image.2022.116633 · Cited by 18
“To evaluate the quality of the reconstructed videos, we used two quality models: the perception-based MS-SSIM [58] and the non-perceptual PSNR.”
MethodSelf-citationAmbiguity of Objective Image Quality Metrics: A New Methodology for Performance Evaluation
Cheon, M., Vigier, T., Krasula, L., Lee, J., Callet, P. L., & Lee, J.-S. (2021) · Signal processing. Image communication · doi:10.1016/j.image.2021.116150 · Cited by 12
“Figure 7 shows the mean ambiguity intervals of two metrics, MS-SSIM and VSNR, for GB of the VDID database with respect to the viewing distance.”
MethodInfluentialCompressive ghost imaging in the presence of environmental noise
Soltanlou, K., & Latifi, H. (2018) · Optics Communications · doi:10.1016/j.optcom.2018.12.024 · Cited by 10
An Information Theoretic Criterion for Image Quality Assessment Based on Natural Scene Statistics
Zhang, D., & Jernigan, E. (2006) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2006.312926 · Cited by 12
GILT: Generating Images from Long TextPreprint
El, O. B., Licht, O., & Yosephian, N. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1901.02404 · Cited by 12
“The most successful method to evaluate image similarity, as mentioned in [7], is multi-scale structural similarity (MS-SSIM [13]).”
BackgroundMethodInfluentialPIRM Challenge on Perceptual Image Enhancement on Smartphones: ReportPreprint
Ignatov, A., Timofte, R., Vu, T., Luu, T. M., Pham, T. X., Nguyen, C. V., Kim, Y., Choi, J.-S., Kim, M., Huang, J., et al. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1810.01641 · Cited by 18
“….pb files that were later run on the test images and validated based on three metrics: • Their speed on HD-resolution (1280 × 720 pixels) images measured compared to the baseline SRCNN [8] network, • PSNR metric measuring their fidelity score, • MS-SSIM [37] metric measuring their perceptual score.”
MethodGraph Signal Processing -- Part I: Graphs, Graph Spectra, and Spectral ClusteringPreprint
Stanković, L., Mandic, D. P., Daković, M., Brajović, M., Dees, B. S., & Constantinides, T. G. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1907.03467 · Cited by 59
“The connectivity weights were calculated using the structural similarity index (SSIM), [69], with an appropriate threshold.”
MethodGenerating Semantic Adversarial Examples via Feature ManipulationPreprint
Wang, S., Nepal, ., Rudolph, C., et al. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2001.02297 · Cited by 10
“To satisfy G3, the structural similarity (SSIM) [22] between the original input and the reconstructed instance should exceed a threshold of θ SIM .”
BackgroundMethodBlind Omnidirectional Image Quality Assessment with Viewport Oriented Graph Convolutional NetworksPreprint
Xu, J., Zhou, W., & Chen, Z. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2002.09140 · Cited by 10
PIPAL: a Large-Scale Image Quality Assessment Dataset for Perceptual Image RestorationPreprint
Gu, J., Cai, H., Chen, H., Ye, X., Ren, J. S. J., & Dong, C. (2020) · European Conference on Computer Vision · doi:10.48550/arxiv.2007.12142 · Cited by 330
“For the FR-IQA methods, we include: PSNR [18], NQM [10], UQI [48], SSIM [49], MS-SSIM [50], IFC [41], VIF [40], VSNR-FR [7], RFSIM [57], GSM [30], SRSIM [55], FSIM and FSIMC [58], SFF [8], VSI [56], SCQI [2], LPIPS-Alex and -VGG [60], PieAPP [37], WaDIQaM [5] and DISTS [11].”
MethodA Cross Channel Context Model for Latents in Deep Image CompressionPreprint
Ma, C., Zhao, W., Liao, R.-L., & Ye, Y. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2103.02884 · Cited by 29
FVC: A New Framework towards Deep Video Compression in Feature SpacePreprint
Hu, Z., Lu, G. S., & Xu, D. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2105.09600 · Cited by 10
“In our proposed framework, we optimize the following Rate Distortion (RD) trade-off: where R o and R r denote the numbers of bits used to encode the offset map O t and the residual feature map R t , respectively. d ( X t , ˆ X t ) denotes the distortion between the input frame X t and the reconstructed frame ˆ X t , w…”
BackgroundMethodInfluentialChen, R., Tang, X., Shen, Z., et al. (2021) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2021.10.08.463746 · Cited by 9
AETHER: Geometric-Aware Unified World Modeling
Zhu, H., Wang, Y., Zhou, J., Chang, W., Zhou, Y., Li, Z., Chen, J., Shen, C., Pang, J., & He, T. (2025) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv51701.2025.00799 · Cited by 133
“Specifically, we introduce three additional loss terms: a Multi-Scale Structure Similarity (MS-SSIM) loss [67] for color video, a scale-and shift-invariant loss [49] for depth videos, and a scale-and shift-invariant pointmap loss [66] for pointmaps projected from the generated depths and raymaps.”
A nonlinear acceleration method for iterative algorithms
Shamsi, M., Ghandi, M., & Marvasti, F. (2019) · Signal Processing · doi:10.1016/j.sigpro.2019.107346 · Cited by 12
Video Compression using Deep Neural Networks
Dayananda, P., Subramanian, S., Suresh, V., et al. (2022) · International Conference on Cognitive Computing and Information Processing · doi:10.1109/ccip57447.2022.10058645 · Cited by 9
“Our deep video codec surpasses all deep learning baselines, MPEG-4 Part 2, and H.261, as measured by MS-SSIM [7] and PSNR, in terms of compression rate and visual quality.”
MethodInfluentialDetecting the Background-Similar Objects in Complex Transportation Scenes
Sun, B., Ma, M., Yuan, N., Li, J., & Yu, T. (2023) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2023.3268378 · Cited by 10
Exploiting Light Polarization for Deep HDR Imaging from a Single Exposure
Pistellato, M., Fatima, T., & Wimmer, M. H. (2023) · Sensors · doi:10.3390/s23125370 · Cited by 9
“We can see that our method produced the highest average PSNR (dB) and MS-SSIM for the test dataset.”
MethodInfluentialHuo, H., Deng, H., Gao, J., Duan, H., & Ma, C. (2023) · Sensors · doi:10.3390/s23156970 · Cited by 10
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.