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 2,701–2,800 of 6,047
RMS-SE-UNet: A Segmentation Method for Tumors in Breast Ultrasound Images
Honghan, Z., Liu, D. C., Jingyan, L., et al. (2021) · IEEE International Conference on Computer and Communication Systems (ICCCS) · doi:10.1109/icccs52626.2021.9449302 · Cited by 8
“Huang et al. proposed a multi-scale structure similarity index(MS-SSIM) [25] loss function to assign higher wights to fuzzy boundary [20].”
BackgroundSaguy, A., Jünger, F., Peleg, A., Ferdman, B., Nehme, E., Rohrbach, A., & Shechtman, Y. (2021) · Optics Express · doi:10.1364/oe.424730 · Cited by 9
Structural similarity analysis of midfacial fractures—a feasibility study
Rozema, R., Kruitbosch, H., van Minnen, B., et al. (2021) · Quantitative Imaging in Medicine and Surgery · doi:10.21037/qims-21-564 · Cited by 8
A Display-Independent Quality Assessment for HDR Images
Zhang, K., Fang, Y., Chen, W., Xu, Y., & Zhao, T. (2022) · IEEE Signal Processing Letters · doi:10.1109/lsp.2022.3141306 · Cited by 9
Stereo Waterdrop Removal with Row-wise Dilated Attention
Shi, Z., Fan, N., Yeung, D., & Chen, Q. (2021) · IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros51168.2021.9636216 · Cited by 13
“To quantitatively evaluate our method, we use the standard PSNR, MS-SSIM [29], and LPIPS [45] as the metrics.”
Fusion of ANNs as decoder of retinal spike trains for scene reconstruction
Li, W., Raj, A. N. J., Tjahjadi, T., & Zhuang, Z. (2022) · Applied Intelligence · doi:10.1007/s10489-022-03402-w · Cited by 9
“03 are two scalar constants [19].”
BackgroundBenchmark of deep learning models for single image super-resolution (SISR)
Soufi, O., Aarab, Z., & Belouadha, F.-Z. (2022) · International Conference on Innovative Research in Applied Science, Engineering and Technology (IRASET) · doi:10.1109/iraset52964.2022.9738274 · Cited by 9
Image denoising using RANSAC and compressive sensing
Stanković, I., Brajović, M., Lerga, J., Daković, M., & Stanković, L. (2022) · Multimedia Tools and Applications · doi:10.1007/s11042-022-13192-5 · Cited by 8
“The MS–SSIM from [53] is a more robust version of the well-known structural similarity (SSIM) index, which is calculated as [52]”
MethodHQ-finGAN: High-Quality Synthetic Fingerprint Generation Using GANs
Sams, A., Shomee, H. H., & Rahman, S. M. M. (2022) · Circuits Systems and Signal Processing · doi:10.1007/s00034-022-02089-1 · Cited by 8
Unsupervised High-Resolution Portrait Gaze Correction and Animation
Zhang, J., Chen, J., Tang, H., Sangineto, E., Wu, P., Yan, Y., Sebe, N., & Wang, W. (2022) · IEEE Transactions on Image Processing · doi:10.1109/tip.2022.3191852 · Cited by 12
Image Compression Using Stochastic-AFD Based Multisignal Sparse Representation
Dai, L., Zhang, L., & Li, H. (2022) · IEEE Transactions on Image Processing · doi:10.1109/tip.2022.3194696 · Cited by 8
Oh, H., Kim, J., Kim, T., & Lee, S. (2022) · IEEE Access · doi:10.1109/access.2022.3209810 · Cited by 10
Two-stage single image Deblurring network based on deblur kernel estimation
Lu, Y., Liu, T. P., & Lin, C. H. (2022) · Multimedia Tools and Applications · doi:10.1007/s11042-022-14116-z · Cited by 9
“[37] have better visual quality but got a lower score in quantitative evaluation.”
BackgroundMethodInfluentialYiu, T. W. H., Fluri, J., & Kacprzak, T. (2022) · Journal of Cosmology and Astroparticle Physics · doi:10.1088/1475-7516/2022/12/013 · Cited by 12
“03 L ) 2 are two small constants, and L is the dynamic range of the pixel values [100].”
BackgroundMethodInfluentialMoniruzzaman, M., Rassau, A., Chai, D., & Islam, S. M. S. (2022) · Journal of Field Robotics · doi:10.1002/rob.22135 · Cited by 11
“According to the literature (Dosselmann & Yang, 2011; Rouse & Hemami, 2008; Søgaard et al., 2016; Z. Wang et al., 2003), MultiScale Structural Similarity Index Measure (MS ‐ SSIM) performs better for different subjective images and video data.”
MethodTCCL-DenseFuse: Infrared and Water Vapor Satellite Image Fusion Model Using Deep Learning
Zhang, C., Guo, J.-X., Ma, L., et al. (2023) · IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing · doi:10.1109/jstars.2023.3277842 · Cited by 8
“The loss function [see (1)] can be used to obtain high-quality fused images while improving the accuracy of TC center positioning where MS_SSIM [22] denotes the multiscale structure similarity index measurement operation, and grad denotes gradient.”
MethodXu, J., Yuan, X., Huang, Y., et al. (2023) · Journal of Biophotonics · doi:10.1002/jbio.202200366 · Cited by 8
Single Image Deraining Using Residual Channel Attention Networks
Wang, D., Pan, J.-S., & Tang, J. (2023) · Journal of Computer Science and Technology · doi:10.1007/s11390-022-0979-2 · Cited by 9
“We evaluate the deraining results on synthetic datasets with PSNR[41] and SSIM[42] metrics.”
MethodA Fast Object Detection-Based Framework for Via Modeling on PCB X-Ray CT Images
Koblah, D. S., Botero, U. J., Costello, S. P., Dizon-Paradis, O. P., Ganji, F., Woodard, D. L., & Forte, D. J. (2023) · ACM Journal on Emerging Technologies in Computing Systems · doi:10.1145/3606948 · Cited by 8
Gu, P., Chen, D. Z., & Wang, C. (2023) · Computers & Graphics · doi:10.1016/j.cag.2023.08.024 · Cited by 13
Evaluation of Image Quality Metrics for Color Prints
Pedersen, M., Zheng, Y., & Hardeberg, J. Y. (2011) · Lecture notes in computer science · doi:10.1007/978-3-642-21227-7_30 · Cited by 8
Exploiting global and local information for image quality assessment with contrast change
Gu, H., Zhai, G., Liu, M., et al. (2015) · IEEE International Symposium on Broadband Multimedia Systems and Broadcasting (BMSB) · doi:10.1109/bmsb.2015.7177190 · Cited by 8
Additive models and separable pooling, a new look at structural similarity
Bondžulić, B. P., & Petrovic, V. S. (2013) · Signal Processing · doi:10.1016/j.sigpro.2013.10.020 · Cited by 9
“SSIM ZW [41] 0.912 0.902 0.377 5.405 MS-SSIM 0.911 0.897 0.382 5.946 AM-SSIM 0.862 0.852 0.467 10.811 AM-GSSIM 0.902 0.894 0.399 7.568 PSNR 0.719 0.688 0.669 20.541 VSNR 0.803 0.799 0.559 15.676 VIF 0.903 0.896 0.410 7.568 Toyama database SSIM 0.798 0.787 0.589 14.286 SSIMlowest 2% 0.847 0.839 0.521 12.500 GSSIM 0.898…”
MethodInfluentialNatural DCT statistics approach to no-reference image quality assessment
Saad, M. A., Bovik, A. C., & Charrier, C. (2010) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2010.5653349 · Cited by 10
“One example is the multi-scale structural similarity index (MS-SSIM) [2] which outperforms the single scale SSIM index.”
MethodSelf-citationA Perceptual Image Quality Assessment Metric Using Singular Value Decomposition
Wang, S., Cui, D., Wang, B., Zhao, B., & Yang, J.-L. (2014) · Circuits, systems, and signal processing · doi:10.1007/s00034-014-9840-3 · Cited by 8
Image quality assessment using Histograms of Oriented Gradients
Yang, Y., Tu, D., & Cheng, G. (2013) · International Conference on Intelligent Control and Information Processing · doi:10.1109/icicip.2013.6568137 · Cited by 9
Robust super-resolution by fusion of interpolated frames for color and grayscale images
Karch, B. K., & Hardie, R. C. (2015) · Frontiers in Physics · doi:10.3389/fphy.2015.00028 · Cited by 8
“To provide alternate quantitative metrics, we also include the Multi-Scale Structural Similarity (MS-SSIM) Index [48].”
MethodAssessment of video naturalness using time-frequency statistics
Mittal, A., Saad, M. A., & Bovik, A. C. (2014) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2014.7025114 · Cited by 9
Self-citationA universal Full Reference image Quality Metric based on a neural fusion approach
Chetouani, A., Beghdadi, A., & Deriche, M. A. (2010) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2010.5652855 · Cited by 9
Study of no-reference image quality assessment algorithms on printed images
Eerola, T., Lensu, L. T., Kalviainen, H. A., & Bovik, A. C. (2014) · Journal of Electronic Imaging · doi:10.1117/1.jei.23.6.061106 · Cited by 9
Self-citationImage quality and QoE in multimedia systems
Klima, M., Fliegel, K., Kekrt, D., et al. (2009) · International Conference Radioelektronika · doi:10.1109/radioelek.2009.5158754 · Cited by 8
Han, Y., Cai, Y., Cao, Y., & Xu, X. (2011) · IEEE Transactions on Image Processing · doi:10.1109/tip.2011.2170697 · Cited by 8
Deterministic techniques for multichannel blind image deconvolution
Souidene, W., Abed‐Meraim, K., & Beghdadi, A. (2006) · International Symposium on Signal Processing and Its Applications (ISSPA) · doi:10.1109/isspa.2005.1580290 · Cited by 8
Rendering of HDR content on LDR displays: an objective approach
Krasula, L., Narwaria, M., Fliegel, K., et al. (2015) · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · doi:10.1117/12.2186388 · Cited by 8
Non-reference assessment of sharpness in blur/noise degraded images
Ospina-Borras, J. E., & Benítez-Restrepo, H. (2016) · Journal of Visual Communication and Image Representation · doi:10.1016/j.jvcir.2016.05.015 · Cited by 9
Mantel, C., Søgaard, J., Bech, S., Korhonen, J., Pedersen, J. M., & Forchhammer, S. O. (2016) · IEEE Transactions on Image Processing · doi:10.1109/tip.2016.2576399 · Cited by 10
“metrics: MSE, PSNR, MS-SSIM [35], HDR-VDP-2 [19] and HDR-VQM [20].”
On Evaluating Perceptual Quality of Online User-Generated Videos
Jang, S., & Lee, J.-S. (2016) · IEEE transactions on multimedia · doi:10.1109/tmm.2016.2581582 · Cited by 8
“Structural similarity (SSIM)[2], multi-scale SSIM (MS-SSIM)[3], most apparent distortion (MAD)[4], and visual information fidelity (VIF)[5] are well-known FR quality metrics for images, and motionbased video integrity evaluation (MOVIE)[6] and spatiotemporal MAD (ST-MAD)[7] are FR metrics for videos.”
MethodShearlet-based reduced reference image quality assessment
Bosse, S., Chen, Q., Siekmann, M., Samek, W., & Wiegand, T. (2016) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2016.7532719 · Cited by 9
Lim, C.-L., & Paramesran, R. (2014) · International Symposium on Intelligent Signal Processing and Communication Systems (ISPACS) · doi:10.1109/ispacs.2014.7024421 · Cited by 8
“In addition to that, we also tabulate the performance of 3 full-reference metrics: peak-signal-to-noise ratio (PSNR), structural similarity index (SSIM) [5], multi-scale structural similarity index (MS-SSIM) [6] and low-level feature similarity index (FSIM and FSIMc) [7].”
BackgroundRaj, S. S., Nair, P. (. ). M. S., & Subrahmanyam, G. R. K. S. (2017) · Journal of the Indian Society of Remote Sensing · doi:10.1007/s12524-016-0654-6 · Cited by 8
CrowdEyes: crowdsourcing for robust real-world mobile eye tracking
Othman, M., Amaral, T., McNaney, R., Smeddinck, J. D., Vines, J., & Olivier, P. L. (2017) · International Conference on Human-Computer Interaction with Mobile Devices and Services (MobileHCI) · doi:10.1145/3098279.3098559 · Cited by 9
“Using the LPW dataset [36] we found the farthest distance between two consecutive frames to be under 15px (Figure 6-left), while under 30px between two consecutive MSSSIM-selected frames (Figure 6-right).”
MethodInfluentialMulti-modal medical image fusion using 2DPCA
Nawaz, Q., Xiao, B., Li, W., et al. (2017) · International Conference on Image, Vision and Computing (ICIVC) · doi:10.1109/icivc.2017.7984635 · Cited by 8
Quality assessment of multi-view-plus-depth images
Wang, J., Wang, S., Zeng, K., & Wang, Z. (2017) · IEEE International Conference on Multimedia and Expo (ICME) · doi:10.1109/icme.2017.8019542 · Cited by 8
Self-citationSegura-García, J., Felici-Castell, S., & García, M. (2018) · Journal of Network and Computer Applications · doi:10.1016/j.jnca.2018.01.006 · Cited by 10
Evidence of change blindness in subjective image fidelity assessment
Moan, S. L., & Pedersen, M. (2017) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2017.8296864 · Cited by 8
A new reduced-reference image quality assessment based on the SVD signal projection
Kalatehjari, E., & Yaghmaee, F. (2018) · Multimedia Tools and Applications · doi:10.1007/s11042-018-5757-3 · Cited by 8
“Therefore, employing the extracted structural features by the SVD tool can lead to a strong correlation between estimated image quality and the HVS’s perception of visual quality.”
BackgroundMethodInfluentialAn error-based video quality assessment method with temporal information
Loh, W.-T., & Bong, D. B. L. (2018) · Multimedia Tools and Applications · doi:10.1007/s11042-018-6107-1 · Cited by 8
“However, there are some differences in relation to Table 4 Pearson linear correlation for the LIVE database Metric Wireless IP H.264 MPEG-2 MSE/PSNR 0.5940 0.3836 0.4112 0.3520 SSIM 0.6653 0.6855 0.7377 0.5980 MSSIM 0.7122 0.7282 0.7341 0.6766 MOVIE [23] 0.8386 0.7622 0.7902 0.7596 ViS3 [35] 0.8574 0.8349 0.7993 0.757…”
MethodInfluentialPerspectives on the definition of visually lossless quality for mobile and large format displays
Allison, R. S., Brunnstrom, K. E., Chandler, D. M., Colett, H. R., Corriveau, P. J., Daly, S. J., Goel, J., Long, J. Y., Wilcox, L. M., Yaacob, Y. M., et al. (2018) · Journal of Electronic Imaging · doi:10.1117/1.jei.27.5.053035 · Cited by 8
“Such metrics are in high demand, given the increasing requirements for real-time image compression needed to deal with the bandwidth requirements of high-resolution image transmission.(22,48) However, it is clear that while subjective testing is labor intensive and costly, it remains the only reliable means of evaluat…”
BackgroundDeep cartoon colorizer: An automatic approach for colorization of vintage cartoons
Chybicki, M., Kozakiewicz, W., Sielski, D., & Fabijańska, A. (2019) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2019.02.006 · Cited by 9
Image quality assessment using a novel region smoothness measure
Khosravi, M., & Hassanpour, H. (2019) · Journal of Visual Communication and Image Representation · doi:10.1016/j.jvcir.2018.11.019 · Cited by 7
Jang, B., Chang, J. H., Park, A. J., & Wu, H. (2019) · Journal of Medical Imaging and Radiation Oncology · doi:10.1111/1754-9485.12868 · Cited by 9
A Sinusoidal-Hyperbolic Family of Transforms With Potential Applications in Compressive Sensing
Abedi, M., Sun, B. S., & Zheng, Z. (2019) · IEEE Transactions on Image Processing · doi:10.1109/tip.2019.2912355 · Cited by 8
Image Quality Assessment of Multi-exposure Image Fusion for Both Static and Dynamic Scenes
Fang, Y., Zeng, Y., Zhu, H., et al. (2019) · IEEE International Conference on Multimedia and Expo (ICME) · doi:10.1109/icme.2019.00083 · Cited by 10
“Since a multi-scale scheme can mimic human visual behavior to explore image details for understanding of image content [18], and human visual system (HVS) is highly sensitive to the gradient and contrast information.”
BackgroundMethodInfluentialShang, X., Zhao, H., Wang, G., Zhao, X., & Zuo, Y. (2019) · IEEE Transactions on Broadcasting · doi:10.1109/tbc.2019.2932286 · Cited by 8
“The peak signal to noise ratio (PSNR) and structure similarity (SSIM) [5], [6] are examples of such schemes.”
BackgroundAdaptive Mesh Texture for Multi-View Appearance Modeling
Armando, M., Franco, J.-S., & Boyer, E. (2019) · International Conference on 3D Vision (3DV) · doi:10.1109/3dv.2019.00082 · Cited by 8
“We use two different metrics for such comparison: the Multi-scale Structural Similarity (MS-SSIM) from [21], and the more recent Learned Perceptual Image Patch Similarity (LPIPS) from [25].”
A Co-Prediction-Based Compression Scheme for Correlated Images
Yin, W., Shi, Y., Zuo, W., et al. (2019) · IEEE Transactions on Multimedia · doi:10.1109/tmm.2019.2949393 · Cited by 8
“The image distortion is evaluated with peak-signal-to-noise-ratio (PSNR) and the multi-scale structure similarity index (MS-SSIM) [28].”
MethodRumora, L., Gašparović, M., Miler, M., & Medak, D. (2019) · International Journal of Image and Data Fusion · doi:10.1080/19479832.2019.1683624 · Cited by 8
“1992), multi-scale structural similarity index (MS-SIM) (Wang et al. 2003) were calculated.”
BackgroundMethodImage quality assessment via spatial‐transformed domains multi‐feature fusion
Yu, M., Zheng, Y., Liao, K., & Tang, Z. (2019) · IET Image Processing · doi:10.1049/iet-ipr.2018.6417 · Cited by 9
A Spectral Hazy Image Database
Khoury, J. E., Thomas, J., & Mansouri, A. (2020) · Lecture notes in computer science · doi:10.1007/978-3-030-51935-3_5 · Cited by 8
“…to evaluate spectral images: PSNR, which calculate the absolute error between images; SSIM [25], which consider the neighborhood dependencies while measuring contrast and structure similarity; and MS-SSIM, which is a multiscale SSIM [26], and performs particularly well in assessing sharpness [7].”
BackgroundAn Extended Hybrid Image Compression Based on Soft-to-Hard Quantification
Fu, H., Liang, F., & Lei, B. (2020) · IEEE Access · doi:10.1109/access.2020.2994393 · Cited by 9
“In the middle of the two downsampling layers, two residual blocks [7] are utilized to further increase the performance of the network.”
Conditional Adversarial Networks for Multimodal Photo-Realistic Point Cloud Rendering
Peters, T., & Claus Brenner, A. (2020) · PFG – Journal of Photogrammetry Remote Sensing and Geoinformation Science · doi:10.1007/s41064-020-00114-z · Cited by 8
High-resolution image de-raining using conditional GAN with sub-pixel upscaling
Sharma, P. K., Basavaraju, S., & Sur, A. (2020) · Multimedia Tools and Applications · doi:10.1007/s11042-020-09642-7 · Cited by 8
“…similarity between de-rained and ground truth images, Peak Signal-to-Noise Ratio (PSNR), Visual Information Fidelity (VIF) [37], Multi-scale Structural Similarity (MS-SSIM) [47], Mean-squared Error (MSE) and Total Variation (TV) to calculate amount of noise present in the image after de-noising.”
BackgroundAccelerating 3D single-molecule localization microscopy using blind sparse inpainting
Gaire, S. K., Wang, Y., Zhang, H. F., et al. (2021) · Journal of Biomedical Optics · doi:10.1117/1.jbo.26.2.026501 · Cited by 8
“To reconstruct the For the quantitative evaluation of the reconstructed images of experimental data, we used the multiscale structural similarity index (MS-SSIM), 57 a perceptually motivated metric, between the reference high-density image and the reconstructed image.”
MethodConstant-Time Gaussian Filtering for Acceleration of Structure Similarity
Sasaki, T., Fukushima, N., Maeda, Y., Sugimoto, K., & Kamata, S. (2020) · International Conference on Information Photonics · doi:10.1109/icip48927.2020.9367337 · Cited by 7
“The multi-scaled version of SSIM, named MS-SSIM [20], is then extended.”
MethodWan, W., Wu, J., Shi, G., Li, Y., & Dong, W. (2018) · IEEE International Conference on Multimedia and Expo (ICME) · doi:10.1109/icme.2018.8486519 · Cited by 8
A blur estimation and detection method for out-of-focus images
Hong, Y., Ren, G., Liu, E., & Sun, J. (2015) · Multimedia Tools and Applications · doi:10.1007/s11042-015-2792-1 · Cited by 8
Automatic scoring of CDMAM using a model of the recognition threshold of the human visual system: R*
Prieto, G., Chevalier, M. L., & Guibelalde, E. (2009) · Proceedings - International Conference on Image Processing · doi:10.1109/icip.2009.5413924 · Cited by 8
Low-Dose CT via Deep CNN with Skip Connection and Network in NetworkPreprint
You, C., Yang, L., Zhang, Y., et al. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1811.10564 · Cited by 8
“e (HU) to the unit interval [0,1] before the images were fed to the network. Finally, we used three common image quality metrics: peak signal-to-noise ratio (PSNR), structural similarity index (SSIM) [18], and root-mean-square error (RMSE) to evaluate the denoised image quality. The visual inspection of our results in…”
MethodDeep SR-ITM: Joint Learning of Super-Resolution and Inverse Tone-Mapping for 4K UHD HDR ApplicationsPreprint
Kim, S. Y., Oh, J., & Kim, M. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1904.11176 · Cited by 8
Progressive Perception-Oriented Network for Single Image Super-ResolutionPreprint
Zheng, H., Li, J., Gao, X., et al. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1907.10399 · Cited by 8
“process of M stages of downsampling, MS-SSIM is defined as MS-SSIM(x;y) = l M (x;y) YM j=1 cs j j (x;y); (6) where l M and cs j are the term we defined in Equation 5 at scale Mand j, respectively. From [31], we set = M and = [0:0448;0:2856;0:3001;0:2363;0:1333]. Therefore, the total loss function of our structure branch…”
MIPGAN -- Generating Robust and High QualityMorph Attacks Using Identity Prior Driven GANPreprint
Zhang, H., Venkatesh, S. K., Ramachandra, R., Raja, K. B., Damer, N., & Busch, C. (2020) · arXiv (Cornell University) · Cited by 9
Unsupervised Anomaly Detection and Localisation with Multi-scale Interpolated Gaussian Descriptors.Preprint
Chen, Y., Tian, Y., Pang, G., & Carneiro, G. (2021) · arXiv (Cornell University) · Cited by 7
Learning from Natural Noise to Denoise Micro-Doppler SpectrogramPreprint
Tang, C., Li, W., Vishwakarma, S., et al. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2102.06887 · Cited by 8
“c1 and c2 are small constants which stabilize the division and more details can be found in [30].”
BackgroundA strong baseline for image and video quality assessmentPreprint
Wen, S., & Wang, J. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2111.07104 · Cited by 32
“The Peak Signal to Noise Ratio (PSNR) [2], the Structural Similarity Index (SSIM) [3] and the Multi-Scale Structural Similarity (MS-SSIM) [4] are usually used as traditional methods for image assessment, but they are not correlate well with human perceptual quality sometimes.”
MethodHyperInverter: Improving StyleGAN Inversion via HypernetworkPreprint
Dinh, T. M., Tran, A. T., Nguyen, R. M. H., et al. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2112.00719 · Cited by 8
“Specifically, we use the pixel-wise L2, perceptual LPIPS [56], MS-SSIM [48], and PSNR metrics.”
BackgroundMethodA Review on Remote Sensing Data Fusion with Generative Adversarial Networks (GAN)Preprint
Liu, P. (2021) · doi:10.36227/techrxiv.16607261 · Cited by 9
“Structural similarity (MS-SSIM) [37] is also a common loss in GAN for RS image fusion.”
BackgroundA Comprehensive Review of Deep Learning-based Single Image Super-resolutionPreprint
Bashir, S. M. A., Wang, Y., Khan, M., et al. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2102.09351 · Cited by 8
Single image super-resolution using capsule generative adversarial network
Hajian, A. (2021) · doi:10.58837/chula.the.2021.128 · Cited by 8
Nam, W., Youn, T., & Ha, C. (2025) · Vehicles · doi:10.3390/vehicles7010008 · Cited by 9
Chen, P., Xu, M., & Wang, X. (2022) · Symmetry · doi:10.3390/sym14122691 · Cited by 7
Complex-valued SAR Image Compression: A Novel Approach for Amplitude and Phase Recovery
Maharjan, P., & Li, Z. (2023) · IEEE International Conference on Visual Communications and Image Processing (VCIP) · doi:10.1109/vcip59821.2023.10402686 · Cited by 7
“We employ the RD-curve as a metric quality evaluation for SAR amplitude image, which plots the bpp against the peak-signal-to-noise-ratio (PSNR) and bpp against the MS-SSIM [28].”
MethodIntelligent visually lossless compression of dental images
Kryvenko, S., Krylova, O., Лукин, В. В., & Kryvenko, S. (2024) · Advanced Optical Technologies · doi:10.3389/aot.2024.1306142 · Cited by 12
Do, H. P., Lockard, C. A., Berkeley, D., Tymkiw, B., Dulude, N., Tashman, S., Gold, G. E., Gross, J. S., Kelly, E., & Ho, C. P. (2024) · Skeletal Radiology · doi:10.1007/s00256-024-04679-3 · Cited by 7
Gerg, I. D., Cook, D. A., & Monga, V. (2024) · IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing · doi:10.1109/jstars.2024.3393139 · Cited by 7
“To measure autofocus efficacy, we use three common image quality assessment (IQA) metrics: peak-signal-to-noise ratio (PSNR) [52], structural similar (SSIM) [52], and multiscale structural similarity (MS-SSIM) [51].”
BackgroundMethodPaired conditional generative adversarial network for highly accelerated liver 4D MRI
Xu, D., Miao, X., Liu, H., Scholey, J., Yang, W., Feng, M. U.-S., Ohliger, M. A., Lin, H., Lao, Y., Yang, Y., et al. (2024) · Physics in Medicine and Biology · doi:10.1088/1361-6560/ad5489 · Cited by 8
“Re-Con-GAN is structured with pair-trained conditional GAN architecture constraint with loss objective fused from L 2 , L 1 and MS-SSIM.”
InfluentialQuality evaluation of point cloud compression techniques
Prazeres, J., Pereira, M., & Pinheiro, A. (2024) · Signal Processing Image Communication · doi:10.1016/j.image.2024.117156 · Cited by 8
Entropy-Based Importance Reordering method for Mitigating Distortion in Slow Fading Channels
Liu, Y., Bao, Z., Liang, H., Zheng, Z., Dong, C., & Xu, X. (2024) · IEEE Wireless Communications and Networking Conference · doi:10.1109/wcnc57260.2024.10571068 · Cited by 7
Generalized Nested Latent Variable Models For Lossy Coding Applied To Wind Turbine Scenarios
Pérez-Gonzalo, R., Espersen, A., & Agudo, A. (2024) · Proceedings - International Conference on Image Processing · doi:10.1109/icip51287.2024.10648110 · Cited by 7
Toward identity preserving in face sketch-photo synthesis using a hybrid CNN-Mamba framework
Tang, D., Jiang, X., Wang, K., Guo, W., Zhang, J., Lin, Y., & Pu, H. (2024) · Scientific Reports · doi:10.1038/s41598-024-72066-y · Cited by 11
“We measured the quality of the generated photos using three metrics: structural similarity image measurement (SSIM) 83 , multiscale SSIM (MS-SSIM) 78 , and Peak signal-to-noise ratio (PSNR) 84 .”
MethodZhang, Y., Fan, R., Duan, P., Dong, J., & Lei, Z. (2024) · IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing · doi:10.1109/jstars.2024.3476153 · Cited by 8
“Many existing methods [10], [11], [41] utilize the perceptuallosscalculatedfromfeaturesextractedbyapretrainednetworktogeneratehigh-qualityimages.”
MethodxU-NetFullSharp: The Novel Deep Learning Architecture for Chest X-ray Bone Shadow Suppression
Schiller, V., Bürget, R., Genzor, S., Mizera, J., & Mezina, A. (2024) · Biomedical Signal Processing and Control · doi:10.1016/j.bspc.2024.106983 · Cited by 8
Dual-stream Perception-driven Blind Quality Assessment for Stereoscopic Omnidirectional Images
Wan, Z., Yang, Q., Li, Z., Fan, X., Zuo, W., & Zhao, D. (2024) · ACM International Conference on Multimedia (ACM MM) · doi:10.1145/3664647.3681561 · Cited by 9
Increasing the Robustness of Image Quality Assessment Models Through Adversarial Training
Chistyakova, A., Antsiferova, A., Khrebtov, M., Lavrushkin, S., Arkhipenko, K. V., Vatolin, D. S., & Турдаков, Д. (2024) · Technologies · doi:10.3390/technologies12110220 · Cited by 11
Deep-learning-based image compression for microscopy images: An empirical study
Zhou, Y., Sollmann, J., & Chen, J. (2024) · Biological Imaging · doi:10.1017/s2633903x24000151 · Cited by 8
“D is the distortion term, which can control the reconstruction quality. ρ is the norm or perceptual metric, e.g. MSE, MS-SSIM [14], etc.”
Halladja, G., Djebbar, A., Boulemden, A., et al. (2025) · Signal Image and Video Processing · doi:10.1007/s11760-025-03816-2 · Cited by 7
“Traditional methods for assessing image similarity typically rely on pixel-level or structural-level comparisons, such as the Peak Signal-to-Noise Ratio (PSNR) [43], Structural Similarity Index (SSIM) [44], and Multiscale Structural Similarity Index (MS-SSIM) [45].”
TCC-SemCom: A Transformer-CNN Complementary Block-Based Image Semantic Communication
Cheng, G., Chong, B., & Lu, H. (2025) · IEEE Communications Letters · doi:10.1109/lcomm.2025.3538486 · Cited by 7
“The performance metrics are pixel-wise metric peak signal-to-noise ratio (PSNR) and perceptual metric multi-scale structural similarity index (MS-SSIM) [23].”
Menakath, M. M., Panicker, M. R., Hareesh, G., et al. (2025) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2025.3545198 · Cited by 7
“The similarity between the sparse array images and the corresponding uniform array images is quantified using the structural similarity index (SSIM) defined in [31].”
Zhao, X., Huang, L., Li, M., Han, C., & Nie, T. (2025) · Remote Sensing · doi:10.3390/rs17122069 · Cited by 9
CGVQM+D: Computer Graphics Video Quality Metric and Dataset
Jindal, A., Sadaka, N. G., Thomas, M. M., Sochenov, A., & Kaplanyan, A. (2025) · Computer Graphics Forum · doi:10.1111/cgf.70221 · Cited by 11
Fast High‐Fidelity Fourier Ptychographic Microscopy via Wavelet Transform and Linear Attention
Sun, W., Li, J., Wu, H., Jiang, X., Zhang, X., Wei, X., Wang, Y., & Zhou, J. (2025) · Laser & Photonics Review · doi:10.1002/lpor.202501895 · Cited by 9
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.