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 4,801–4,900 of 6,047
Hybrid Security System based on Wavelet Domain Watermarking and Chaotic Map Cryptography
Al-qdah, M. (2015) · International Journal of Computer Applications · doi:10.5120/19373-9190 · Cited by 1
Mobile Video Quality Assessment: A Current Challenge for Combined Metrics
Okarma, K. (2014) · Advances in intelligent systems and computing · doi:10.1007/978-3-319-06740-7_41 · Cited by 1
Using structural information for reduced reference image quality assessment
Kalatehjari, E., & Yaghmaee, F. (2014) · International Conference on Computer and Knowledge Engineering · doi:10.1109/iccke.2014.6993368 · Cited by 1
“Some of the FR-IQAs are such as MSE (Mean Squared Error), visual signal-tonoise ratio (VSNR) [1] and MS-SSIM (Multi-Scale Structural Similarity) [2].”
MethodInfluentialSuresh, H., Hegde, S., & Sartori, J. (2017) · doi:10.1145/3139315.3139319 · Cited by 1
“MS-SSIM has been shown to align more closely with subjective evaluations of image quality than SSIM and PSNR [36].”
MethodSupervised dictionary learning for blind image quality assessment
Jiang, Q., Shao, F., Jiang, G., et al. (2015) · Visual Communications and Image Processing · doi:10.1109/vcip.2015.7457897 · Cited by 1
Quality Metric for Shadow Rendering
Kluczek, K. (2016) · Annals of Computer Science and Information Systems · doi:10.15439/2016f494 · Cited by 1
Image quality assessment based on Structure Similarity
Wu, J., Li, H., & Xia, Z. (2016) · IEEE International Conference on Signal Processing, Communications and Computing (ICSPCC) · doi:10.1109/icspcc.2016.7753620 · Cited by 1
Boosting in image quality assessment
Temel, D., & AlRegib, G. (2016) · IEEE International Workshop on Multimedia Signal Processing (MMSP) · doi:10.1109/mmsp.2016.7813335 · Cited by 1
Image Quality Assessment Scheme Based on Structural Contrast Index and Gradient Similarity
Liu, L., Zheng, Y., & Wang, W. (2017) · Lecture notes in electrical engineering · doi:10.1007/978-981-10-3530-2_41 · Cited by 1
Comparison of Some Image Quality Approaches.
Parfenenkov, B. B., & Panachev, M. A. (2014) · Cited by 1
Quality Assessment of High-Resolution Images with Small Distortions after Compression
Petkus, T., Tichonov, J., Filatovas, E., & Jakštys, V. (2017) · Baltic Journal of Modern Computing · doi:10.22364/bjmc.2017.5.2.04 · Cited by 1
“Despite the variety of the SSIM-based methods, in this paper, we focus on the investigation of the original SSIM index method (Wang et al., 2004a), which is the basis of the others.”
MethodInfluentialSaliency Based Assessment of Videos from Frame-Wise Quality Measures
Roja, B., & Sandhya, B. (2017) · IEEE International Advance Computing Conference · doi:10.1109/iacc.2017.0135 · Cited by 1
“2) Multi scale Structural Similarity Index (MSSIM): The multi scale structural similarity measure [18] is a more flexible method than the other single scale methods such as SSIM.”
MethodWu, X., Sun, H., Ma, X., Wang, Y., & Han, Y. (2017) · Infrared Physics & Technology · doi:10.1016/j.infrared.2017.07.011 · Cited by 2
Application of full-reference video quality metrics in IPTV
Sedano, I., Prieto, G., Brunnstrom, K. E., Kihl, M., & Montalbán, J. (2017) · IEEE International Symposium on Broadband Multimedia Systems and Broadcasting (BMSB) · doi:10.1109/bmsb.2017.7986191 · Cited by 1
A new method with saliency detection for image quality assessment
Xiao, R.-Y., & Yeh, M.-H. (2017) · IEEE International Conference on Consumer Electronics - Taiwan (ICCE-Taiwan) · doi:10.1109/icce-china.2017.7991002 · Cited by 2
“2 we show the scatter plots of subjective mean opinion scores (MOS) against objective scores predicted by some representative IQA indices (FSIMc, FSIM, MS-SSIM, PSNR-HA, PSNR-HMA, VSNR, PSNR-HVS, PSNR-HVSM, and SDSIM)on TID2013.”
MethodLearning from neighbours for saliency optimization
Lin, W., Niu, Y., Lin, J., et al. (2017) · 2017 Computing Conference · doi:10.1109/sai.2017.8252278 · Cited by 1
“In order to evaluate the proposed LFN optimization algorithm, this paper used a variety of saliency evaluation metrics , namely the mean squared error (MSE), the multi-scale structural similarity index method (MS-SSIM) [11], the peak signal-to-noise ratio based on the human visual system (PSNR-HVS) [12], the modified…”
MethodInfluentialHow to Optimize the Utilization of Image Quality Metrics in Computer Vision?
Chetouani, A., & Hassouni, M. E. (2018) · Mediterranean Conference on Pattern Recognition and Artificial Intelligence (MedPRAI) · doi:10.1145/3177148.3180097 · Cited by 1
“In order to better explicit it, Table I shows a ranking of 12 common used FR measures (VIFP / VIF [9], WSNR [10], PSNRM [11], PSNRH [12], SSIM [13], UQI [14], NQM [15], IFC [16], XYZ [17], VSNR [18]).”
MethodEfficient and effective objective image quality assessment metrics
Nafchi, H. Z. (2018) · Cited by 1
Luo, X., Lai, W. W., & Chang, R. K. W. (2018) · International Conference on Ground Penetrating Radar (GPR) · doi:10.1109/icgpr.2018.8441553 · Cited by 1
Computational sensing approaches for enhanced active imaging
Laurenzis, M. (2018) · doi:10.1117/12.2325566 · Cited by 2
Image Up-Sampling for Super Resolution with Generative Adversarial Network
Tsunekawa, S., Inoue, K., & Yoshioka, M. (2018) · Lecture notes in computer science · doi:10.1007/978-3-030-03991-2_26 · Cited by 1
Stereoscopic Display Technologies and Their Applications in Medical Imaging
Albani, L., Marchessoux, C., & Kimpe, T. R. L. (2011) · Information Display · doi:10.1002/j.2637-496x.2011.tb00435.x · Cited by 1
No Reference Quality Assessment of Blurred Images
Baig, M. A., Moinuddin, A. A., & Khan, E. (2018) · IEEE Uttar Pradesh Section International Conference on Electrical, Electronics and Computer Engineering (UPCON) · doi:10.1109/upcon.2018.8596837 · Cited by 1
“However, other popular variants of SSIM such as multi-scale SSIM (MS-SSIM) [9] and information weighted SSIM (IWSSIM) [11] are also available.”
MethodObjective Quality Assessment of 2D Synthesized Views for Light-Field Visualization
Tamboli, R. R., Kara, P. A., Bisht, N., et al. (2018) · International Conference on 3D Immersion (IC3D) · doi:10.1109/ic3d.2018.8657904 · Cited by 1
“As seen in Figure 4d, the values of 3 structural-similarity metrics — namely SSIM, MS-SSIM and FSIM — saturate for higher skip values.”
MethodFast Implementation of Image Structural Similarity Algorithm
Xu, Y., Zhuang, Z., Li, W., & Fanga, Y. (2019) · IOP Conference Series Materials Science and Engineering · doi:10.1088/1757-899x/533/1/012057 · Cited by 2
Influence of Different Descriptors to Enhance Image Registration Techniques Using FREAK: Case Study
Aarathi, M. R., & Raju, J. (2019) · doi:10.1109/iciict1.2019.8741354 · Cited by 1
Quality Index for Benchmarking Image Inpainting Algorithms with Guided Regional Statistics
Song, L., Li, L., Hu, B., & ZHANG, J. (2019) · IEICE Transactions on Information and Systems · doi:10.1587/transinf.2018edl8206 · Cited by 1
Incorporating Gradient Direction for Assessing Multiple Distortions
Cheraaqee, P., Mansouri, A., & Mahmoudi-Aznaveh, A. (2019) · doi:10.1109/pria.2019.8785992 · Cited by 1
“The rival methods are PSNR, VSNR [7], SSIM [4], VIF [6], FSIM [8], VSI [9], MS-SSIM [5], SR-SIM [10], UQI [3] and GMSD [14].”
MethodA Correlation-Based No-Reference Packet-Loss Metric
Morais, D. D. R., da Silva, A. F., & Farias, M. (2016) · Anais de XXXIV Simpósio Brasileiro de Telecomunicações · doi:10.14209/sbrt.2016.148 · Cited by 1
A Learning Based Image Quality Assessment Model Assisted with Visual Saliency and Gradient Features
Liu, H., Yang, H., Pan, Z., Huang, B., & Wang, J. (2019) · IEEE International Conference on Signal and Image Processing (ICSIP) · doi:10.1109/siprocess.2019.8868544 · Cited by 1
“The algorithm we proposed was compared with other 8 representative IQA models and the CNN-IQA model, including PSNR, IFC, SSIM, G-SSIM, MS-SSIM, GMSD, VSI, FSIM.”
MethodInfluentialiTM-Net: Deep Inverse Tone Mapping Using Novel Loss Function Based on Tone Mapping Operator
Kinoshita, Y., & Kiya, H. (2019) · European Signal Processing Conference (EUSIPCO) · doi:10.23919/eusipco.2019.8902744 · Cited by 1
“2 [26], and PU encoding [27] with MS-SSIM [28] which 2019 27th European Signal Processing Conference (EUSIPCO)”
BackgroundInfluentialHierarchical Learning-Guided human motion quality assessment in big data environment
Huang, Z., Li, Y., & Luo, S. (2019) · Journal of Visual Communication and Image Representation · doi:10.1016/j.jvcir.2019.102700 · Cited by 1
Perceptual visual quality assessment using deeply-learned gaze shifting kernel
Zhao, F., Huang, S., Long, R., et al. (2019) · Journal of Visual Communication and Image Representation · doi:10.1016/j.jvcir.2019.102701 · Cited by 1
High Quality Network and Device Aware Multimedia Content Delivery for Mobile Cloud
Saleem, M., & Saleem, Y. (2019) · KSII Transactions on Internet and Information Systems · doi:10.3837/tiis.2019.10.004 · Cited by 1
“SSIM extracts the structural information using HVS and structural similarity is measured between the original and the image under observation [33].”
MethodBlind image quality assessment based on statistics features and perceptual features
Zhao, Y., Ji, X., & Liu, Z. (2019) · Journal of Intelligent & Fuzzy Systems · doi:10.3233/jifs-190998 · Cited by 1
Neural Adaptive Transport Framework for Internet-scale Interactive Media Streaming Services
Xu, Y., Chen, H., Zhao, Y., et al. (2019) · IEEE International Symposium on Broadband Multimedia Systems and Broadcasting (BMSB) · doi:10.1109/bmsb47279.2019.8971955 · Cited by 1
“It is reported about 30% bitrate reduction for the same quality evaluated using the multiscale structural similarity (MS-SSIM) [12], [13].”
Aguilar, S. E. P., Mújica‐Vargas, D., & Kinani, J. M. V. (2018) · Research in Computing Science · doi:10.13053/rcs-147-7-12 · Cited by 1
“El tercero, es el ´Indice de Similitud Estructural (SSIM) [13], ver ecuaci´on (12), es una medida cuantitativa de la diferencia entre la imagen original y reconstruida en cuanto a sus luminancias, contrastes e informaci´on de estructura.”
BackgroundEfficiency Estimation of the Inter-Frame Coding of Modern Video Codecs
Gavrilov, I. A., Nosirov, K., Puziy, A., et al. (2019) · International Conference on Information Science and Communications Technologies (ICISCT) · doi:10.1109/icisct47635.2019.9011871 · Cited by 1
“Therefore, SEP metric shows more fair values, than MSE and PSNR [3, 5, 6].”
BackgroundReconstructing 3D Human Avatars from Monocular Images
Alldieck, T., Kappel, M., Castillo, S., & Magnor, M. A. (2020) · Lecture notes in computer science · doi:10.1007/978-3-030-41816-8_8 · Cited by 1
“We maximize the structural similarity of ground truth and predicted normal and displacement maps by minimizing the dissimilarity (MS-DSSIM): (1−MS-SSIM)/2.”
MethodMunich to Dubai: How far is it for Semantic Segmentationƒ
Rai, S. N., Balasubramanian, V. N., Subramanian, A., et al. (2020) · IEEE Winter Conference on Applications of Computer Vision · doi:10.1109/wacv45572.2020.9093400 · Cited by 1
“Additionally, we use Multi-Scale Structural Similarity (MS-SSIM) [34] to measure the structural similarity between images at various resolutions.”
MethodGuo, J.-M., Hsia, C., Wu, C.-S., et al. (2020) · IEEE International Conference on Consumer Electronics - Taiwan (ICCE-Taiwan) · doi:10.1109/icce-taiwan49838.2020.9258234 · Cited by 1
Li, C., Shen, X., & Li, H. (2020) · IEEE Access · doi:10.1109/access.2020.3040758 · Cited by 1
“The Structural Similarity (SSIM) [33] can compare the essential differences between the generated images and the real images.”
BackgroundMSLp: Deep Superresolution for Meteorological Satellite Image
Zhao, L., Yu, H., & Wang, Y. (2021) · Complexity · doi:10.1155/2021/2678124 · Cited by 1
Recent Advancements in Medical Imaging: A Machine Learning Approach
Dang, N., Tiwari, S., Khurana, M., & Arya, K. V. (2021) · Studies in big data · doi:10.1007/978-981-15-9492-2_10 · Cited by 2
Towards Artifacts-Free Image Defogging
Graffieti, G., & Maltoni, D. (2021) · Proceedings - International Conference on Pattern Recognition/Proceedings/International Conference on Pattern Recognition · doi:10.1109/icpr48806.2021.9412289 · Cited by 1
“In fact, even if improved metrics have been introduced (such as MS-SSIM [28]), SSIM and PSNR have been used by most of the methods we considered for comparison.”
MethodDeep Neural Networks for Ring Artifacts Segmentation and Corrections in Fragments of CT Images
Kornilov, A., Safonov, I. V., Reimers, I., & Yakimchuk, I. (2021) · Conference of Open Innovations Association (FRUCT) · doi:10.23919/fruct50888.2021.9347587 · Cited by 4
“The valuable loss functions for image restoration with neural networks [36] are the weighted sum of mean absolute error (MAE) and multiscale structural similarity index measure (MS-SSIM) [37].”
BackgroundAn Efficient Video Compression Network
Dhungel, P., Tandan, P., Bhusal, S., et al. (2020) · International Conference on Advances in Computing, Communication Control and Networking (ICAC3N) · doi:10.1109/icacccn51052.2020.9362823 · Cited by 1
DarkGAN: Night Image Enhancement Using Generative Adversarial Networks
Alaspure, P., Hambarde, P., Dudhane, A. A., & Murala, S. (2021) · Communications in computer and information science · doi:10.1007/978-981-16-1086-8_26 · Cited by 3
Cross-Content Evaluations in the Subjective Quality Assessment of Light Field Images
Zizien, A., & Fliegel, K. (2021) · International Conference Radioelektronika · doi:10.1109/radioelektronika52220.2021.9420203 · Cited by 1
“What is more, given the results of the presented subjective evaluation, together with the results introduced in LFDD [5] and by Paudyal et al. in [18], commonly used metrics, such as PSNR and MS-SSIM [23], present a viable option for objective quality evaluation of light field content.”
ResultHow Do Different Tasks Influence Each Other’s Learning? Case Study In Dermoscopic Images
Talavera-Martínez, L., Bibiloni, P., & González-Hidalgo, M. (2021) · IEEE International Symposium on Biomedical Imaging (ISBI) · doi:10.1109/isbi48211.2021.9433814 · Cited by 1
Design of Low-Artifact Interpolation Kernels by Means of Computer Algebra
Karpov, P. (2022) · Mathematics in Computer Science · doi:10.1007/s11786-022-00538-3 · Cited by 1
“Since the root mean square error (RMSE) is not a reliable indicator of perceived image quality, we employed 12 additional full-reference image quality assessment (IQA) methods implemented in the PIQ library [14]: SSIM [37], MSSSIM [27], VIFp [24], FSIM [33], GMSD [29], VSI [31], HaarPSI [21], MDSI [38], MS-GMSD [30],…”
MethodMulti-core Median Redescending M-Estimator for Impulsive Denoising in Color Images
Mújica‐Vargas, D., Rendón-Castro, A., Matuz-Cruz, M., et al. (2021) · Lecture notes in computer science · doi:10.1007/978-3-030-77004-4_25 · Cited by 1
Similarity and symmetry measures based on fuzzy descriptors of image objects’ composition
Iwanowski, M., & Grza̧bka, M. (2021) · Computer Science Research Notes · doi:10.24132/csrn.2021.3002.33 · Cited by 1
Stereoscopic image quality assessment considering visual mechanism and multi-loss constraints
Li, S., Li, Y., & Han, Y. (2021) · Journal of Visual Communication and Image Representation · doi:10.1016/j.jvcir.2021.103255 · Cited by 1
An, Y., Du, X., & Martins, J. R. R. A. (2021) · AIAA AVIATION Forum · doi:10.2514/6.2021-3061 · Cited by 3
Acceleration by Deep-Learnt Sharing of Superfluous Information in Multi-contrast MRI
Chatterjee, S., Joel, S. E., Venkatesan, R., et al. (2021) · Lecture notes in computer science · doi:10.1007/978-3-030-87231-1_38 · Cited by 1
Valdés, J. J., & Pou, A. (2021) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn52387.2021.9533823 · Cited by 1
“The state-of-the-art MSSIM similarity measure [19], [20], is conceptually a combination of three components characterizing the luminance ( l ), contrast ( c ) and structure of the image ( s ).”
BackgroundCase Studies to Demonstrate Real-World Applications in Ophthalmic Image Analysis
Remeseiro, B., & Bolón‐Canedo, V. (2021) · Intelligent systems reference library · doi:10.1007/978-3-030-79161-2_4 · Cited by 1
Mechanism Analysis and Mitigation of Visual Navigation System Vulnerability
Zhai, Y., Fu, Y., Wang, S., & Zhan, X. (2021) · Lecture notes in electrical engineering · doi:10.1007/978-981-16-3142-9_49 · Cited by 1
Fine-Tuning of the Measure for Full Reference Image Quality Assessment
Gorokhovatskyi, O., & Peredrii, O. (2021) · Lecture notes on data engineering and communications technologies · doi:10.1007/978-3-030-82014-5_29 · Cited by 1
Deep Convolutional Neural Network for Compressive Sensing of Magnetic Resonance Images
Lu, H., Zou, X., Liao, L., Li, K., & Liu, J. (2021) · International Journal of Pattern Recognition and Artificial Intelligence · doi:10.1142/s0218001421520194 · Cited by 1
Image quality assessment based on manifold distortion
Türkan, M. (2021) · Pamukkale University Journal of Engineering Sciences · doi:10.5505/pajes.2020.69158 · Cited by 1
Mudeng, V., Kim, M., & Choe, S. (2021) · Biosensors · doi:10.3390/bios11120504 · Cited by 1
“The multiscale SSIM (MS-SSIM) may be more flexible than the mean SSIM (MSSIM) because it provides multiscale image assessment with downsampling by two images in each iteration [29,30].”
MethodBaldo, D., Renzone, G. D., Fort, A., et al. (2021) · ACTA IMEKO · doi:10.21014/acta_imeko.v10i4.1093 · Cited by 1
“The assessment of the variation of picture quality was carried out by resorting to the Multiscale Structural Similarity (MSSSIM) index for image quality [25].”
MethodGPU-Intensive Fast Entropy Coding Framework for Neural Image Compression
Akutsu, H., Naruko, T., & Suzuki, A. (2021) · IEEE International Conference on Visual Communications and Image Processing (VCIP) · doi:10.1109/vcip53242.2021.9675405 · Cited by 1
“All the components save the entropy coder are composed of differentiable functions, so neural image compression has the unique advantage of being able to optimize differentiable quality indexes, e.g., multi-scale structural similarity (MS-SSIM [14]) and GANs [5] [3].”
BackgroundMethodEfficient Local Image Descriptors Learned With Autoencoders
Žižakić, N., & Pižurica, A. (2021) · IEEE Access · doi:10.1109/access.2021.3138168 · Cited by 5
“For more details on MS-SSIM, refer to [42].”
BackgroundDeep learning enables accelerated optical coherence tomography angiography
Kim, G., Kim, J., Choi, W. J., et al. (2022) · doi:10.1117/12.2609303 · Cited by 1
“The results indicate that our DNN is superior to the baseline method in terms of image reconstruction measures (i.e., SSIM, MS-SSIM, and PSNR) and angiogram quality measures (i.e., the contrast and vessel connectivity) for all downsampling ratios.”
MethodEnhanced optical flow-based full reference video quality assessment algorithm
Gujjunoori, S., Oruganti, M., & Pais, A. R. (2022) · Multimedia Tools and Applications · doi:10.1007/s11042-022-12591-y · Cited by 2
“To obtain the spatial distortion features the MS-SSIM index [26] is used.”
MethodDe-Speckling of SAR Images with Fuzzy Filters along with Altered Preserved Edge Values
Mynoddin, M., Kabir, M. F., Islam, M. N., et al. (2022) · Journal of Computer and Communications · doi:10.4236/jcc.2022.103002 · Cited by 1
“The higher value of SSIM is the best value for the output image [29].”
BackgroundHair color digitization through imaging and deep inverse graphics
Kips, R., Bokaris, P., Perrot, M., et al. (2022) · Electronic Imaging · doi:10.2352/ei.2022.34.10.ipas-383 · Cited by 1
Artificial Intelligence for Image Enhancement and Reconstruction in Magnetic Resonance Imaging
Hammernik, K., & Akçakaya, M. (2022) · Contemporary medical imaging · doi:10.1007/978-3-030-92087-6_13 · Cited by 3
Sezer, V., Öncü, S., & Baykas, P. B. (2020) · IntechOpen eBooks · doi:10.5772/intechopen.87712 · Cited by 2
Performance Analysis of Face Detection system using HOG and QualHOG Features
(2019) · International Journal of Engineering and Advanced Technology · doi:10.35940/ijeat.f8474.088619 · Cited by 1
Mitigating Object Prior-Bias From Sparse-Projection Tomographic Reconstructions
Gopal, P., Chandran, S., Svalbe, I., & Rajwade, A. V. (2022) · IEEE Transactions on Computational Imaging · doi:10.1109/tci.2022.3172589 · Cited by 1
Xiao, J., & Zhang, D. (2022) · International Conference on Image, Vision and Computing (ICIVC) · doi:10.1109/icivc55077.2022.9886420 · Cited by 2
Bilevel Methods for Image Reconstruction
Crockett, C., & Fessler, J. A. (2022) · doi:10.1561/9781638280033 · Cited by 4
Hong, Y., Wang, C., & Jiang, X. (2022) · Electronics · doi:10.3390/electronics11193155 · Cited by 1
Transforming Image Generation from Scene Graphs
Sortino, R., Palazzo, S., & Spampinato, C. (2022) · Proceedings - International Conference on Pattern Recognition/Proceedings/International Conference on Pattern Recognition · doi:10.1109/icpr56361.2022.9956620 · Cited by 4
“We also compute MS-SSIM [42] to give an evaluation on the image structure, and FID [43] to assess coherence between generated and ground-truth image features.”
MethodLearning to regulate 3D head shape by removing occluding hair from in-the-wild images
Anisetty, S., Saravanabavan, V., & Yiyu, C. (2022) · IEEE International Symposium on Mixed and Augmented Reality Adjunct (ISMAR-Adjunct) · doi:10.1109/ismar-adjunct57072.2022.00087 · Cited by 1
“To further improve the quality of the inversion we introduce the M-SSIM loss [34] which additionally gives a quantifiable metric to measure inversion accuracy.”
MethodMutual information based objective model for assessment of visual quality
Mohanty, B., & Sahoo, T. (2022) · Journal of Information and Optimization Sciences · doi:10.1080/02522667.2022.2094552 · Cited by 1
A Light Field Image Quality Assessment Method Based on Stereo Vision
Fu, W., Shen, X., Zhou, W., et al. (2022) · IEEE International Conference on Unmanned Systems (ICUS) · doi:10.1109/icus55513.2022.9986647 · Cited by 1
“At present, many FR algorithms have been raised in FR 2D-IQA, In SSIM (Structural Similarity Index) [3], which weighs the Structural Similarity between reference and distorted images, several variations of it have also been proposed.”
MethodAttention-based Image Compression in Sensor Assembly
Meier, S., Erkan, A., Thielen, N., et al. (2022) · IEEE International Symposium for Design and Technology in Electronic Packaging (SIITME) · doi:10.1109/siitme56728.2022.9987988 · Cited by 1
“Reference [20] takes the SSIM algorithm to evaluate image quality using a sliding window approach, where the window moves pixel by pixel across the entire image space.”
MethodMSCI: A Multi-Source Composite Image Database for Compression Distortion Quality Assessment
Zhang, X., Xu, Z., Lin, Z., & Wang, M. (2022) · IEEE International Conference on Visual Communications and Image Processing (VCIP) · doi:10.1109/vcip56404.2022.10008864 · Cited by 1
“For full-reference, the selected methods are MSSSIM [42], VIF [24], SQMS [20], SVQI [21], GFM [22] and FGCQA [23].”
MethodReversing Image Signal Processors by Reverse Style Transferring
Kınlı, F., Özcan, B., & Kıraç, F. (2023) · Lecture notes in computer science · doi:10.1007/978-3-031-25063-7_43 · Cited by 2
Low-Complexity Quality Measurement for Real-Time Video Compression
Cock, J. D., Decker, A. D., & Sivashanmugam, S. (2022) · SMPTE Media Technology Summit · doi:10.5594/m001958 · Cited by 1
Patch-Based Image Learned Codec using Overlapping
Tarchouli, M., Rivière, M., Guionnet, T., Hamidouche, W., Outtas, M., & Olivier Déforges, A. (2023) · Signal & Image Processing An International Journal · doi:10.5121/sipij.2023.14101 · Cited by 1
“The loss function to be minimized is the ratedistortion loss function formulated as : 𝐽 = 𝐷 + 𝜆𝑅, (2) where D refers to the distortion between the original patch and the reconstructed one, measured by the Mean Square Error (MSE) or the Multi-Scale Structural Similarity Index [25] (MS-SSIM) metrics, and R refers to the…”
MethodHYBRID SCHEMES BASED ON WAVELET TRANSFORM AND CONVOLUTIONAL AUTO-ENCODER FOR IMAGE COMPRESSION
Chakib, H., Idrissi, N., & Jannani, O. (2023) · International Journal of Engineering Science Technologies · doi:10.29121/ijoest.v7.i2.2023.479 · Cited by 1
Song, J., Chang, Y., Park, S., et al. (2023) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp49357.2023.10095189 · Cited by 1
“In the light of this, we employed six metrics that can shed light on the different aspects of the quality of reconstruction; PSNR, SSIM [32], MS-SSIM [33], RMSE, LPIPS [43], and FID [10].”
MethodA just noticeable distortion based perceptually lossless image compression codec
Liu, K. (2012) · OPUS Publication Server of the University of Stuttgart (University of Stuttgart) · doi:10.18419/opus-2894 · Cited by 1
Assessment of Quality of 3D Displays
Lueder, E. H. (2011) · doi:10.1002/9781119962762.ch4 · Cited by 1
A distortion-based ranking of image quality indices using mutual information
Deriche, M. A., Beghdadi, A., & Chetouani, A. (2010) · International Conference on Information Science, Signal Processing and their Applications (ISSPA) · doi:10.1109/isspa.2010.5605450 · Cited by 1
Multi-scale and Scalable Video Quality Assessment
Seshadrinathan, K., & Bovik, A. C. (2008) · 2008 Second International Conference on Electrical Engineering · doi:10.1109/icee.2008.4585297 · Cited by 1
Self-citationColorization for Image Compression.Preprint
Baig, M. H., & Torresani, L. (2016) · arXiv (Cornell University) · Cited by 1
Learning Adaptive Parameter Tuning for Image ProcessingPreprint
Dong, J., Frosio, I., & Kautz, J. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1610.09414 · Cited by 1
“We used two different cost functions: the mean PSNR on the training set, e = =1 PSNR (˜ o j , o j ) ; and the mean MS-SSIM [21], e = =1 MS-SSIM (˜ o j , o j ) .”
BackgroundMethodBlind quality assessment of image and speech signals
Li, Q. (2017) · DR-NTU (Nanyang Technological University) · doi:10.32657/10356/70758 · Cited by 1
“In addition, both the statistical structural and luminance features are calculated at three scales to account for the variations of viewing distance and image resolution [10, 76].”
BackgroundMethodRestricting Greed in Training of Generative Adversarial NetworkPreprint
You, H., Jiao, Z., Xu, H., Li, J., Wang, Y., & Gao, X. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1711.10152 · Cited by 1
“Independent Wasserstein critic [16], inception score [17] and multi-scale structural similarity [18] are adopted in our experiments.”
MethodColour image segmentation using perceptual colour difference saliency algorithm
Bukola, T. T. (2017) · DUT Open Scholar (Durban University of Technology) · doi:10.51415/10321/2931 · Cited by 1
DeepSIC: Deep Semantic Image CompressionPreprint
Luo, S., Yang, Y., & Song, M. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1801.09468 · Cited by 1
“Metric To assess the visual quality of reconstructed images, we adopt Multi-Scale Structural Similarity Index Metric (MS-SSIM) [34] and Peak Signal-to-Noise Ratio (PSNR) [8] for comparing original, uncompressed images to compressed, degraded ones.”
MethodVideo quality assessment using subjective and objective metrics
Gaurav, R. (2013) · Akademik Açık Erişim (Işık Üniversitesi) · Cited by 1
Contrast and visual saliency similarity-induced index for assessing image qualityPreprint
Jia, H., Zhang, L., & Wang, T. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1708.06616 · Cited by 1
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.