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 3,301–3,400 of 6,047
PERCEPTUAL QUALITY ASSESSMENT OF DENOISED IMAGES
Zeng, K., & Wang, Z. (2015) · Cited by 7
Self-citationKhan, M. I., Jeoti, V., & Malik, A. S. (2010) · International Conference on Computer Applications and Industrial Electronics · doi:10.1109/iccaie.2010.5735022 · Cited by 5
Revisiting Perceptual Distortion for Natural Images: Mean Discrete Structural Similarity Index
Hillar, C. J., & Marzen, S. (2017) · DCC · doi:10.1109/dcc.2017.84 · Cited by 6
“MSSIM and variants thereof [4, 5] have been used widely, in some cases replacing MSE/PSNR, although the structural component in MSSIM’s computation is directly related to a local (patch-wise) MSE [6, 7].”
MethodBlind video quality assessment based on spatio-temporal internal generative mechanism
Zhu, Y. H., Wang, Y., & Shuai, Y. (2017) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2017.8296292 · Cited by 6
“From Table1 and Table2, it is obvious that our proposed model outperforms the FR-VQA algorithms MS-SSIM and VQM, as these methods are based on the image quality assessment, which couldn’t catch the temporal distortions.”
MethodInfluentialRedefined Block-Lifting-Based Filter Banks With Efficient Reversible Nonexpansive Convolution
Suzuki, T., Tanaka, N., & Kudo, H. (2018) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2018.2849101 · Cited by 5
“) where BIT and MSE are the number of bits of the image and the mean squared error, and multi-scale structural similarity (MS-SSIM) [35], which is an image quality assessment based on human visual characteristics unlike PSNR, in lossy image coding.”
MethodFan, Y., Larabi, M., Cheikh, F. A., & Fernández-Maloigne, C. (2018) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2018.8451490 · Cited by 5
“First, multi-scale LoG- and GM-based features will be considered to improve the performance of our method because of the usefulness of multiple scales for IQA [29].”
MethodSRQM: A Video Quality Metric for Spatial Resolution Adaptation
Mackin, A., Afonso, M., Zhang, F., & Bull, D. (2018) · Picture Coding Symposium · doi:10.1109/pcs.2018.8456246 · Cited by 5
“The performance of SRQM is compared against five quality metrics: PSNR [27], SSIM [24], VSNR [28], MS-SSIM [29] and VIF [30].”
MethodFlexible Motion Vector Resolution Prediction for Video Coding
Ray, B., Larabi, M., & Jung, J. (2018) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2018.2878401 · Cited by 5
Characterization of the sources of degradation in remote sensing satellite images
Hussein, M. S., Hanafy, M. E., & Mahmoud, T. A. (2019) · doi:10.1109/itce.2019.8646604 · Cited by 5
Objective Quality Evaluation of an Angularly-Continuous Light-Field Format
Tamboli, R. R., Cserkaszky, A., Kara, P. A., Barsi, A., & Martini, M. G. (2018) · International Conference on 3D Immersion (IC3D) · doi:10.1109/ic3d.2018.8657876 · Cited by 5
“‘MS-SSIM’ was used to find similarity between the two optical flow arrays.”
MethodInfluentialA multi-metric approach for block-level video quality assessment
Papadopoulos, M. A., Katsenou, A., Agrafiotis, D. K., et al. (2019) · Signal Processing Image Communication · doi:10.1016/j.image.2019.06.009 · Cited by 5
“…for images, such as the Structural Similarity Index (SSIM) [1], Peak Signal to Noise Ratio based on HVS (PSNRHVSM) [3], Multi-Scale SSIM (MS-SSIM) [4], Visual Information Fidelity (VIF) [5], Feature Similarity Index (FSIM) [6]; or for video, such as Perception-based Video Metric (PVM) [7],…”
MethodA new end-to-end image compression system based on convolutional neural networks
Akyazi, P., & Ebrahimi, T. (2019) · SPIE Applications of Digital Image Processing · doi:10.1117/12.2530195 · Cited by 7
Zuo, F., Gao, J., Yang, D., et al. (2019) · IEEE Intelligent Transportation Systems Conference (ITSC) · doi:10.1109/itsc.2019.8917034 · Cited by 5
Siamese Score: Detecting Mode Collapse for GANs
Jia, J., & Zhao, Q. (2019) · International Congress on Image and Signal Processing (CISP) · doi:10.1109/cisp-bmei48845.2019.8965997 · Cited by 5
Codedretrieval: Joint Image Compression and Retrieval with Neural Networks
Liu, L., Liu, H., Chen, T., Shen, Q., & Ma, Z. (2019) · IEEE International Conference on Visual Communications and Image Processing (VCIP) · doi:10.1109/vcip47243.2019.8965918 · Cited by 5
“For image compression, a MS-SSIM loss [13] is em ployed to improve the reconstruction quality:”
MethodScalable double regularization for 3D Nano-CT reconstruction
Tang, W., & Li, M. (2020) · Journal of Petroleum Science and Engineering · doi:10.1016/j.petrol.2020.107271 · Cited by 5
“noise ratio (SNR) and the structural similarity (SSIM) index [45] between the reconstructed structure f̂ and the true structure f :”
BackgroundDCM-CNN: Densely Connected Multiloss Convolutional Neural Networks for Light Field View Synthesis
Liu, L., Nian, Z., & Jung, C. (2020) · IEEE Access · doi:10.1109/access.2020.2989453 · Cited by 6
An Anomaly Detection Method for Outdoors Insulator in High-Speed Railway Traction Substation
Lu, X., Peng, Y., Quan, W., Zhou, N., Zou, D., & Chen, J. X. (2020) · International Conference on Advances in Computer Technology, Information Science and Communications (CTISC) · doi:10.1109/ctisc49998.2020.00032 · Cited by 6
“Multi-scale structural similarity and anomaly detection The multi-scale structure similarity algorithm MS-SSIM [10] is used to extract the feature differences between the input image and the reconstructed image.”
MethodTrain Small, Generate Big: Synthesis of Colorectal Cancer Histology Images
Deshpande, S., Minhas, F., & Rajpoot, N. (2020) · Lecture notes in computer science · doi:10.1007/978-3-030-59520-3_17 · Cited by 6
Bonsai Style Classification: a new database and baseline results
Nakahata, G. H. D. S., Constantino, A. A., & Costa, Y. M. G. (2020) · IEEE International Symposium on Multimedia (ISM) · doi:10.1109/ism.2020.00025 · Cited by 5
Ho, M. M., & Zhou, J. (2022) · IEEE Winter Conference on Applications of Computer Vision · doi:10.1109/wacv51458.2022.00097 · Cited by 5
“All comparisons are qualitative and quantitative conducted on DIV2K-SCAN valset and testset described in Section 2.4 using similarity metrics such as Peak Signalto-Noise Ratio (PSNR), LPIPS [43], and MS-SSIM [38].”
BackgroundMethodInfluentialKarthik, K., Kamath, S. S., & Kamath, S. U. (2020) · IEEE International Conference on Industrial and Information Systems (ICIIS) · doi:10.1109/iciis51140.2020.9342715 · Cited by 5
“c) Multi-scale Structural Similarity Measure: MS-SSIM [16] is a variant of SSIM where the process of SSIM is applied over multiple scales through sub-sampling.”
BackgroundMethodSparse flow adversarial model for robust image compression
Zhao, S., Yang, S., Liu, Z., Feng, Z., & Zhang, K. (2021) · Knowledge-Based Systems · doi:10.1016/j.knosys.2021.107284 · Cited by 6
MOVI-Codec: Deep Video Compression without Motion
Chen, M., Patney, A., & Bovik, A. C. (2021) · Picture Coding Symposium · doi:10.1109/pcs50896.2021.9477463 · Cited by 7
Self-citationCrack Detection Method of Sleeper Based on Cascade Convolutional Neural Network
Li, L., Zheng, S., Wang, C., Zhao, S., Chai, X., Peng, L., Tong, Q., & Ji, W. (2022) · Journal of Advanced Transportation · doi:10.1155/2022/7851562 · Cited by 6
“…of each loss. k is taken as hybrid loss function l ( k ) that mixes three losses of BCE, SSIM, and IOU is used to obtain a high-quality detection object with complete information: where l ( k ) bce , l ( k ) ssim , and l ( k ) iou denote the BCE [37], SSIM [38], and IOU losses [39], respectively.”
Klyuchnikov, N., Ismailova, L., Kovalev, D., Safonov, S., & Koroteev, D. A. (2022) · Journal of Petroleum Science and Engineering · doi:10.1016/j.petrol.2022.110106 · Cited by 8
A 1Mb Mixed-Precision Quantized Encoder for Image Classification and Patch-Based Compression
Nguyen, V. T., Guicquéro, W., & Sicard, G. (2022) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2022.3145024 · Cited by 5
Gvozdev, O. G., Maiorov, A. A., & Materuhin, A. V. (2022) · Geodesy and Cartography · doi:10.22389/0016-7126-2021-978-12-23-33 · Cited by 5
Automated image quality appraisal through partial least squares discriminant analysis
Ramani, R. G., & Shanthamalar, J. J. (2022) · International Journal of Computer Assisted Radiology and Surgery · doi:10.1007/s11548-022-02668-2 · Cited by 5
MEFuse: end-to-end infrared and visible image fusion method based on multibranch encoder
Hong, Y., Wu, X., & Xu, T. (2022) · Journal of Electronic Imaging · doi:10.1117/1.jei.31.3.033043 · Cited by 5
Ji, Z., Song, X., Feng, Q., et al. (2022) · IEEE Transactions on Intelligent Transportation Systems · doi:10.1109/tits.2022.3176222 · Cited by 5
“MS-SSIM [14] measures the difference between ground-truth images and inpainted ones in terms of luminance, contrast and structure.”
BackgroundMethodInfluentialImproving Image Compression With Adjacent Attention and Refinement Block
Jeny, A. A., Islam, M. B., Junayed, M. S., & Das, D. (2022) · IEEE Access · doi:10.1109/access.2022.3195295 · Cited by 10
“The distortion measure (MS-SSIM [33]) is denoted by d(.).”
MethodRegion-of-Interest based Video Coding Strategy for Low Bitrate Surveillance Systems
Aliouat, A., Kouadria, N., Maimour, M., & Harize, S. (2022) · International Multi-Conference on Systems, Signals & Devices (SSD) · doi:10.1109/ssd54932.2022.9955963 · Cited by 8
“We calculate the PSNR, SSIM, SM-SSIM [33] and VIF [34] between the original frame and a reconstructed frame.”
MethodImpact of different loss functions on denoising of microscopic images
Shah, Z. H., Muller, M. W., Hammer, B., et al. (2022) · 2022 International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn55064.2022.9892936 · Cited by 5
“To examine and evaluate all of the above-mentioned loss functions fairly, we kept all other parameters constant in the training process.”
BackgroundMethodA neural video codec with spatial rate-distortion control
Fathima, N., Petersen, J., Sautière, G., Wiggers, A., & Pourreza, R. (2023) · IEEE Winter Conference on Applications of Computer Vision · doi:10.1109/wacv56688.2023.00533 · Cited by 12
“Other wellestablished handcrafted metrics include Structural Similarity Index Measure (SSIM) [44] and its Multi-Scale variant (MS-SSIM) [45].”
BackgroundPredictive Sampling for Efficient Pairwise Subjective Image Quality Assessment
Mohammadi, S., & Ascenso, J. (2023) · ACM International Conference on Multimedia (ACM MM) · doi:10.1145/3581783.3612443 · Cited by 5
“MS-SSIM [21] Boosts SSIM metric by considering the variations in image resolution and viewing conditions.”
MethodLow-Latency Neural Stereo Streaming
Hou, Q., Farhadzadeh, F., Said, A., Sautière, G., & Le, H. (2024) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52733.2024.00762 · Cited by 5
“When reporting MS-SSIM performance, we use a version of our network which is further finetuned using MS-SSIM as distortion loss for an additional 100k iterations.”
MethodA Novel PSNR-B Approach for Evaluating the Quality of De-blocked Images
Tadala, T. (2012) · IOSR Journal of Computer Engineering · doi:10.9790/0661-0454049 · Cited by 5
Modeling sparse spatio-temporal representations for no-reference video quality assessment
Shabeer, P. M., Bhati, S., & Channappayya, S. S. (2017) · IEEE Global Conference on Signal and Information Processing · doi:10.1109/globalsip.2017.8309155 · Cited by 5
“1 shows the frame-wise MS-SSIM index [13] values of the reconstructed frames across different dictionary sizes.”
BackgroundMethodFull-reference image quality assessment based on the analysis of distortion process
Ma, X., Jiang, X., & Guo, X. (2017) · International Conference on Systems and Informatics (ICSAI) · doi:10.1109/icsai.2017.8248471 · Cited by 6
“Meanwhile, some metrics evaluate image quality by measuring the structural similarity, such as structural similarity index (SSIM) [11], multi-scale SSIM (MS-SSIM) [12] and feature similarity index (FSIM) [13].”
BackgroundSaliency inspired quality assessment of stereoscopic 3D video
Banitalebi-Dehkordi, A., & Nasiopoulos, P. (2018) · Multimedia Tools and Applications · doi:10.1007/s11042-018-5837-4 · Cited by 5
“In order to apply saliency prediction to MS-SSIM, we generate the saliency maps at each scale independently.”
BackgroundMethodInfluentialSmooth Deep Image Generator from Noises
Guo, T., Xu, C., Shi, B., Xu, C., & Tao, D. (2019) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v33i01.33013731 · Cited by 6
“Multi-scale Structural Similarity for Image Quality (MS-SSIM) (Wang, Simoncelli, and Bovik 2003) was proposed to measuring the similarity of two images.”
MethodInfluentialA Spatial-Temporal Weighted Method for Asymmetrically Distorted Stereo Video Quality Assessment
Fang, Y., Sui, X., & Wang, J. (2019) · IEEE International Symposium on Circuits and Systems proceedings · doi:10.1109/iscas.2019.8702796 · Cited by 5
“Corresponding to the first stage of the proposed method, we use four 2D-IQA methods (PSNR, SSIM, MS-SSIM and IW-SSIM) and one 2D-VQA methods (VQM) to estimate the 2D video quality of single-view.”
3-D Visual Discomfort Assessment Considering Optical and Neural Attention Models
Yang, J., Nguyen, V., Sim, K., Zhao, Y. S., & Lu, W. B. (2019) · IEEE Transactions on Broadcasting · doi:10.1109/tbc.2019.2932293 · Cited by 5
A No-Reference Autoencoder Video Quality Metric
Martinez, H. B., Farias, M. C. Q., & Hines, A. (2019) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2019.8803204 · Cited by 5
“The FR quality metrics are SSIM [17] and PSNR, while the NR metrics are VIIDEO [18], DIIVINE [12], BIQI [19], NIQE [20], and BRISQUE [21].”
MethodQuality Assessment for Natural and Screen Visual Contents
Loh, W.-T., & Bong, D. B. L. (2019) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2019.8803254 · Cited by 6
“The natural image QA methods involved are Peak Signal to Noise Ratio (PSNR) [11], SSIM, Multiscale SSIM (MSSIM) [12], Information Weighted SSIM (IWSSIM) [13], feature similarity (FSIM) [14], GMSD, and visual information fidelity (VIF) [15].”
MethodXuan, K., Wei, D., Wu, D., et al. (2019) · Lecture notes in computer science · doi:10.1007/978-3-030-32248-9_8 · Cited by 5
Blind Prediction of Original Image Quality for Sentinel Sar Data
Rubel, O. S., Rubel, A., Лукин, В. В., Carli, M., & Egiazarian, K. (2019) · European Workshop on Visual Information Processing · doi:10.1109/euvip47703.2019.8946231 · Cited by 5
Qiu, M., & Zhu, Y. (2020) · PLoS ONE · doi:10.1371/journal.pone.0231617 · Cited by 5
“The PSNR, SSIM, MS-SSIM, FSIM results of NEDI and RBC are referenced from [26].”
MethodInfluentialRobust HDR image quality assessment using combination of quality metrics
Choudhury, A. (2020) · Multimedia Tools and Applications · doi:10.1007/s11042-020-08985-5 · Cited by 5
“On database 4 [52], although HDR-CQM is better, it is statistically equivalent to existing HDR methods and the best performing SDR (MS-SSIM) method.”
BackgroundMethodInfluentialZeeshan, M., & Majid, M. (2020) · Telecommunication Systems · doi:10.1007/s11235-020-00702-9 · Cited by 5
“Objective evaluation of our proposed PMDVC scheme was performed using multi-scale structural similarity index (MS-SSIM) [46].”
MethodVisual stream connectivity predicts assessments of image quality
Bowen, E. F. W., Rodriguez, A. M., Sowinski, D., & Granger, R. (2022) · Journal of Vision · doi:10.1167/jov.22.11.4 · Cited by 5
“We compare the predictive capacity of approaches I and II against performance-driven IQA measures from the literature: SSIM, MS-SSIM (Wang et al., 2003), IWSSIM (Wang & Li, 2011), VSNR (Chandler & Hemami, 2007), VIF (Sheikh & Bovik, 2006), VIFP (Sheikh & Bovik, 2006), IFC (Sheikh, Bovik, & De…”
Multi-Mode Intra Prediction for Learning-Based Image Compression
Jung, H. C., Guerin, N. D., Ramos, R. S., et al. (2020) · Proceedings - International Conference on Image Processing · doi:10.1109/icip40778.2020.9191108 · Cited by 5
“Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index Measure (SSIM) [20], and Multi Scale SSIM (MSSSIM) [21] quality scores are depicted in Fig.”
MethodAmirpour, H., Pinheiro, A., Fonseca, E., et al. (2020) · Proceedings - International Conference on Image Processing · doi:10.1109/icip40778.2020.9190663 · Cited by 5
“The five selected metrics were PSNR, SSIM [13], MSSSIM [14], FSIM [15], and VIFP [16].”
MethodLi, Y., & Chen, R. (2020) · IET Image Processing · doi:10.1049/iet-ipr.2019.1677 · Cited by 5
“In this section, we use the clear images as ground truth for SSIM, which is a reference index.”
MethodInfluentialArtifact-Free Single Image Defogging
Graffieti, G., & Maltoni, D. (2021) · Atmosphere · doi:10.3390/atmos12050577 · Cited by 5
“In fact, even if improved versions of such metrics have been introduced (such as MS-SSIM [41]), SSIM and PSNR have been used by most of the methods we considered for comparison, and they remain two of the most used metrics in the defogging literature.”
MethodChen, L., Han, J., & Tian, F. (2021) · Journal of Intelligent & Fuzzy Systems · doi:10.3233/jifs-210987 · Cited by 5
“In order to evaluate the perceptual quality we adopt noise quality measure (NQM) [39], structure similarity index (SSIM) [40] and multiscale structure similarity index (MSSIM) [41] as the quantitative evaluation metrics.”
High Frame Rate Video Quality Assessment using VMAF and Entropic Differences
Madhusudana, P. C., Birkbeck, N., Wang, Y., Adsumilli, B., & Bovik, A. C. (2021) · Picture Coding Symposium · doi:10.1109/pcs50896.2021.9477462 · Cited by 9
“Compared Methods : We choose 3 FR-IQA methods: PSNR, SSIM [6] and MS-SSIM [7], and 3 FR-VQA models: ST-RRED [12], SpEED [21] and VMAF [19] for performance comparison.”
MethodSelf-citationNo-Reference Image Quality Assessment Based on Edge Pattern Feature in the Spatial Domain
Shao, W., & Mou, X. (2021) · IEEE Access · doi:10.1109/access.2021.3112741 · Cited by 6
A Convolutional Sparsity Regularization for Solving Inverse Scattering Problems
Song, R., Zhou, Q., Liu, Y., Li, C., & Chen, X. (2021) · IEEE Antennas and Wireless Propagation Letters · doi:10.1109/lawp.2021.3108040 · Cited by 5
“The use of CSR makes the reconstruction results of SOM-CSR more robust against noise compared to SOM, which can be clearly observed from the SSIM indices.”
MethodInfluentialReduction of Spatially Correlated Speckle in Textured SAR Images
Rubel, O. S., Лукин, В. В., Krivenko, S., et al. (2021) · International Journal of Computing · doi:10.47839/ijc.20.3.2276 · Cited by 5
Deep learning colorful ptychographic iterative engine lens-less diffraction microscopy
Bian, Y., Jiang, Y., Wang, J., et al. (2021) · Optics and Lasers in Engineering · doi:10.1016/j.optlaseng.2021.106843 · Cited by 5
Efficient Immersive Video Compression using Screen Content Coding
Samelak, J., Dziembowski, A., Mieloch, D., et al. (2021) · Computer Science Research Notes · doi:10.24132/csrn.2021.3101.22 · Cited by 5
Fast multiscale directional filter bank-based speckle mitigation in gallstone ultrasound images
Leavline, E. J., Sutha, S., & Singh, D. A. A. G. (2014) · Journal of the Optical Society of America A · doi:10.1364/josaa.31.000283 · Cited by 5
Combination of SSIM and JND with content-transition classification for image quality assessment
Hsu, M.-C., Wu, G., & Chien, S. (2012) · Visual Communications and Image Processing · doi:10.1109/vcip.2012.6410840 · Cited by 6
Calibrating MS-SSIM for compression distortions using MLDS
Charrier, C., Knoblauch, K., Maloney, L. T., et al. (2011) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2011.6116381 · Cited by 5
“Among well-known IQA algorithms, Multi-Scale Structural SIMilarity (MS-SSIM) [2] computes relative quality scores between a reference image and a distorted version, achieving excellent correlations with MOS values.”
BackgroundMethodSelf-citationNonlinearities in Stereoscopic Phase-Differencing
Monaco, J. P., Bovik, A. C., & Cormack, L. K. (2008) · IEEE Transactions on Image Processing · doi:10.1109/tip.2008.2001405 · Cited by 5
Self-citationReduced reference watermark-based image transmission quality metric
Kayvanrad, M. H., Sodagari, S., Avanaki, A. N., & Ahmadi-Noubari, H. (2008) · International Symposium on Communications, Control and Signal Processing · doi:10.1109/isccsp.2008.4537282 · Cited by 5
“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 [ 12 ].”
BackgroundStatistical Modeling of Image Degradation Based on Quality Metrics
Chetouani, A., Beghdadi, A., & Deriche, M. A. (2010) · Proceedings - International Conference on Pattern Recognition/Proceedings/International Conference on Pattern Recognition · doi:10.1109/icpr.2010.180 · Cited by 7
Image Quality Assessment Incorporating the Interaction of Spatial and Spectral Sensitivities of HVS
Hu, A., Zhang, R., Zhan, X., & Dong, Y. (2011) · Signal and Image Processing · doi:10.2316/p.2011.759-017 · Cited by 5
Convolutional Sparse Coding for Capturing High‐Speed Video Content
Serrano, A., Garcés, E., Masiá, B., et al. (2017) · Computer Graphics Forum · doi:10.1111/cgf.13086 · Cited by 5
A linear regression framework for assessing time-varying subjective quality in HTTP streaming
Eswara, N., Dendi, S. V. R., Chakraborty, S., Sethuram, H. P., Kuchi, K., Kumar, A., & Channappayya, S. S. (2017) · IEEE Global Conference on Signal and Information Processing · doi:10.1109/globalsip.2017.8308598 · Cited by 6
“MS-SSIM [19] 0.”
No-reference image quality assessment based on quality patches in real time
Zhang, C., Yang, X., Huang, X., Yu, G., & Chen, S. (2018) · EURASIP Journal on Image and Video Processing · doi:10.1186/s13640-018-0361-z · Cited by 5
“[14], which evaluated image quality by using structural information as image quality features, according to the structural sensitiveness of the human vision system.”
MethodImage Quality Assessment Based on Improved Structural SIMilarity
Wu, J., Qi, F., & Shi, G. (2012) · Lecture notes in computer science · doi:10.1007/978-3-642-34778-8_14 · Cited by 5
“1333 from l = 1 to l = M , respectively, according to [10].”
BackgroundMethodHigh definition video quality assessment metric built upon full reference ratios
Jiménez, D. (2012) · UPM Digital Archive (Technical University of Madrid) · doi:10.20868/upm.thesis.14712 · Cited by 6
“In addition, linear rescaling was applied to the SSIM [80], MS-SSIM [83], VSSIM [9], and VIF [120] metrics to facilitate numerical convergence of the nonlinear regression.”
MethodReview of Existing Objective QoE Methodologies
Fang, Y., Lin, W., & WINKLER, S. S. (2015) · Wiley eBooks · doi:10.1002/9781118736135.ch3 · Cited by 6
Editorial: Diversidad y multiculturalidad
Abbas, S., Hussain, M. Z., & Irshad, M. (2016) · Ánfora · doi:10.18187/pjsor.v12i4.1459 · Cited by 5
Vertolli, M. O., & Davies, J. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1708.02237 · Cited by 9
“Bottom-up approaches suffer from a number of known problems despite a number of improvements that they have made in quantifying the error signal based on psychophysical experiments (Wang et al. 2004; Wang and Bovik 2006).”
BackgroundMethodLearning Content-Weighted Deep Image CompressionPreprint
Li, M., Zuo, W., Gu, S., et al. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1904.00664 · Cited by 5
Dense Pose TransferPreprint
Neverova, N., Güler, R. A., & Kokkinos, I. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1809.01995 · Cited by 13
FontGAN: A Unified Generative Framework for Chinese Character Stylization and De-stylizationPreprint
Liu, X., Meng, G., Xiang, S., & Pan, C. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1910.12604 · Cited by 10
“We use Multi-Scale Structural Similarity (MS-SSIM) [21], Local Distortion (LD) [23], L1 loss and OCR accuracy as the evaluation metrics.”
MethodAsymmetric GANs for Image-to-Image TranslationPreprint
Tang, H., & Sebe, N. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1912.06931 · Cited by 8
“To make the model benefit from multi-scale deep information, we refer to a multi-scale implementation of SSIM [31] which constrains SSIM over 𝑀 scales, Domain Classification Loss.”
MethodActive Fine-Tuning From gMAD Examples Improves Blind Image Quality AssessmentPreprint
Wang, Z., & Ma, K.-D. (2020) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.48550/arxiv.2003.03849 · Cited by 43
“In addition to testing the model generalizability using [19] and deepIQA [18], when competing with a full-reference IQA method, MS-SSIM [28], in the gMAD competition on the Waterloo Exploration Database [29]. gMAD, here we shift our attention to leveraging gMAD examples to improve the BIQA…”
MethodSelf-citationLearning Context-Based Non-local Entropy Modeling for Image CompressionPreprint
Li, M., Zhang, K., Zuo, W., et al. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2005.04661 · Cited by 5
“Two separate metrics, i . e ., standard mean square error (MSE) and the perceptual metric MS-SSIM [42], are adopted as the distortion loss.”
MethodDeep Implicit Volume CompressionPreprint
Tang, D., Singh, S., Chou, P. A., et al. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2005.08877 · Cited by 5
“[67] and MS-SSIM [66]), our method preserves better spatio-temporal coherence and thus has better compression rate.”
MethodReference Guided Face Component EditingPreprint
Deng, Q., Cao, J., Liu, Y.-F., Chai, Z.-H., Li, Q., & Sun, Z.-N. (2020) · International Joint Conference on Artificial Intelligence · doi:10.48550/arxiv.2006.02051 · Cited by 22
“…methods [He et al., 2019; Wu et al., 2019], we leverage Fréchet Inception Distance (FID, lower value indicates better quality) [Heusel et al., 2017] and Multi-Scale Structural SIMilarity (MS-SSIM, higher value indicates better quality) [Wang et al., 2003] to evaluate the performance of our model.”
MethodBidirectional Mapping Generative Adversarial Networks for Brain MR to PET SynthesisPreprint
Hu, S., Lei, B., Wang, Y., Feng, Z., Shen, Y., & Wang, S. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2008.03483 · Cited by 5
“where l M and cs j are the terms we defined in SSIM at scale M and j , respectively. α and β j are the constants used to adjust the relative importance of different components, according to [39].”
MethodImproving Deep Video Compression by Resolution-adaptive Flow CodingPreprint
Hu, Z., Chen, Z., Xu, D., et al. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2009.05982 · Cited by 5
“D = d(Xt, X̂t) denotes the distortion between the input frame and the reconstructed frame, where d(·) represents the metric (mean square error or MS-SSIM [31]) for measuring the difference between two images.”
BackgroundMethodSeo, S., Ki, S., & Kim, M. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1902.05316 · Cited by 5
Deep Learning-based Face Super-Resolution: A SurveyPreprint
Jiang, J., Wang, C., Liu, X., et al. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2101.03749 · Cited by 5
“Thus, multi-scale structural similarity index measure (MS-SSIM) [43] is proposed, which divides the image into multiple windows, first assesses SSIM for every window separately, and then converges them to obtain MS-SSIM.”
MethodTian, Y., Pang, G., Liu, F., et al. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2103.03423 · Cited by 5
“Building upon the pre-trained encoder fθ(·) using the loss in (1), we fine-tune two state-of-the-art UAD methods, IGD [8] and F-anoGAN [37], and a baseline method, multi-scale structural similarity index measure (MS-SSIM)-based auto-encoder [48].”
BackgroundMethodInfluentialPerceptual Image Quality Assessment with TransformersPreprint
Cheon, M., Yoon, S., Kang, B., et al. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2104.14730 · Cited by 5
“For each distorted image, predicted scores of PSNR, SSIM [51], MS-SSIM [52], LPIPS [61], and proposed IQT are listed.”
BackgroundMethodInfluentialOn the relation between statistical learning and perceptual distancesPreprint
Hepburn, A., Laparra, V., Santos‐Rodríguez, R., Ballé, J., & Malo, J. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2106.04427 · Cited by 9
“The illustrations consider (a) two representative measures of perceptual distortion: the Normalized Laplacian Perceptual Distance (the NLPD [11]) and the classical Multi-Scale SSIM [8], and (b) a recent model for the probability of natural images, the PixelCNN++ [21]1.”
MethodLearning Foveated Reconstruction to Preserve Perceived Image Statistics.Preprint
Surace, L., Wernikowski, M., Tursun, O. T., Myszkowski, K., Mantiuk, R., & Didyk, P. (2021) · arXiv (Cornell University) · Cited by 5
Sikka, A., Peri, S. V., Virk, J. S., Niyaz, U., & Bathula, D. R. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2108.02160 · Cited by 27
“Hence, [35] proposed a multi-scale variant known as multi-scale structural similarity index (MSSSIM) that computes SSIM at multiple scales.”
Versatile Learned Video CompressionPreprint
Feng, R., Guo, Z., Zhang, Z., et al. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2111.03386 · Cited by 5
“Evaluation Setting We measure the quality of reconstructed frames using PSNR and MS-SSIM [Wang et al. 2003] in the RGB colorspace.”
InfluentialEnd-to-End Learned Image Compression with Quantized Weights and ActivationsPreprint
Sun, H., Yu, L., & Katto, J. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2111.09348 · Cited by 11
“When optimizing MS-SSIM, λ is picked up from {3, 5, 10, 40, 80, 128}.”
BackgroundInfluentialPT-VTON: an Image-Based Virtual Try-On Network with Progressive Pose Attention TransferPreprint
Zhou, H., Lan, T., & Venkataramani, G. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2111.12167 · Cited by 5
“Table 1 shows a comparison of performance between our work against CP-VTON [19], SwapNet [20] and SieveNet [17] on image quality metrics benchmark, IS [21] for image quality, SSIM [22] and MS-SSIM [23] for pair-wise structural similarity.”
MethodSurvey Paper: Image Quality AssessmentPreprint
Choudhary, R. R., Goel, V., & Meena, G. (2019) · SSRN Electronic Journal · doi:10.2139/ssrn.3356307 · Cited by 7
A Weakly Supervised Deep Generative Model for Complex Image Restoration and Style TransformationPreprint
Dai, W., Wong, I. H. M., & Wong, T. T. W. (2022) · doi:10.36227/techrxiv.19747057.v1 · Cited by 6
Metadata
- Authors
- Wang, Z., Simoncelli, E. P., & Bovik, A. C.
- Affiliations
- Courant Institute of Mathematical Sciences (US)
- The University of Texas at Austin (US)
- Venue
- The Thrity-Seventh Asilomar Conference on Signals, Systems & Computers, 2003
- Publisher
- IEEE
- Published
- 2004-07-08
- Pages
- 1398-1402
- ISSN
- 1058-6393 (ISSN-L)
- OpenAlex
- W1580389772
- Cited by
- 6,253
- References
- 15
- Citations by year
- 2026: 344 · 2025: 675 · 2024: 729 · 2023: 668 · 2022: 662 · 2021: 641 · 2020: 563 · 2019: 438 · 2018: 305 · 2017: 267 · 2016: 206 · 2015: 173 · 2014: 166 · 2013: 112 · 2012: 89
- FWCI
- 3.98
- Type
- conference paper
- Language
- en
- Open access
- closed
- Locations
- Asilomar Conference on Signals, Systems and Computers (ACSSC) (publishedVersion)
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Image and Video Quality Assessment
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Image Retrieval and Classification Techniques
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Digital Media Forensic Detection
- Keywords
- structural similarity, image quality assessment, multiscale analysis, human visual system, perceptual image quality, image generation, parameter calibration, viewing conditions
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
A language model drew the claims from the paper on 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.