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 5,601–5,700 of 6,047
Research on Intelligent Traffic Surveillance Video Compression Quality Assessment Method
Zhao, X.-N., Wu, T., & Shen, Q. (2025) · Communications in computer and information science · doi:10.1007/978-981-96-6310-1_10
Cheong, S. Y., Mustafa, A., & Gilbert, A. (2025) · Lecture notes in computer science · doi:10.1007/978-3-031-91569-7_17
A Dual-Stage Image Enhancement Approach Using Weighted Clahe and CnnsPreprint
Priya, S. (2025) · SSRN Electronic Journal · doi:10.2139/ssrn.5270885
CD-DVC: Conditional Diffusion Model for Distributed Video Coding
Ye, Z., Huang, Q., & Tang, H. (2025) · Asia Conference on Algorithms, Computing and Machine Learning (CACML) · doi:10.1109/cacml64929.2025.11010930
“Therefore, this chapter comprehensively adopts PSNR, MS-SSIM [32] and LPIPS [33] as evaluation indicators.”
Xu, B., & Guo, S. (2025) · Quality Engineering · doi:10.1080/08982112.2025.2509008
“Although HVS-based methods (Mannos and Sakrison 1974) can improve perceptual accuracy, their complexity limits practical application (Wang, Simoncelli, and Bovik 2003).”
Guo, Z., Wu, X., & Xie, D. (2025) · Geophysical Journal International · doi:10.1093/gji/ggaf200 · Cited by 1
Prototypical Distribution Divergence Loss for Image Restoration
Peng, J., Fu, J., & Liu, D. (2025) · IEEE Transactions on Image Processing · doi:10.1109/tip.2025.3572818
“Apart from perceptual losses, Zhao et al. [11] suggested employing the MS-SSIM metric [15] as an auxiliary loss to enhance structural consistency.”
Imai, T., Mizutani, T., Iguchi, T., & Haneda, T. (2025) · Report · doi:10.2749/tokyo.2025.3200
Joint channel–spatial entropy modeling for efficient visual coding
Li, Y., Jiang, X., & Sun, Z. (2025) · Neural Computing and Applications · doi:10.1007/s00521-025-11138-0
End-to-End Design of Webtoon-Style Portrait Stylization System for Real-World Demo Booth
Kim, S., Sohn, B.-S., & Lee, J. (2025) · IEEE Access · doi:10.1109/access.2025.3578868
“We used three metrics for d ( · ): Learned Perceptual Image Patch Similarity (LPIPS) [33], Multiscale structural Similarity (MS-SSIM) loss [34], and Cosine similarity loss (CSIM), which capture perceptual similarity, structural consistency, and semantic identity preservation, respectively.”
Privacy-preserving synthetic fingerprint generation with pore-level details
Dhaneshwar, R., Kaur, M., & Kaur, M. (2025) · Multimedia Tools and Applications · doi:10.1007/s11042-025-20978-w
“The higher the score per batch, the lower the amount of diversity among the generated images [55].”
Zhao, M., Wang, S., Xiao, F., Zhang, J., Cheng, X., & Zhu, Y. (2025) · International Conference on Computer Supported Cooperative Work in Design (CSCWD) · doi:10.1109/cscwd64889.2025.11033287
SaLIC: Saliency-Enhanced Learned Image Compression for Balanced Quality
He, M., Zou, J., Xu, Y., Song-lin, S., & Wang, J. (2025) · IEEE International Symposium on Circuits and Systems proceedings · doi:10.1109/iscas56072.2025.11044220
“Model performance was evaluated on Kodak [37], DIV2K [38], CLIC2020 [42], and CLIC2024 [27] datasets using PSNR, MS-SSIM [28], LPIPS [25], and FID [26].”
The visibility of Eidolon distortions in things and stuff
Mahncke, S., Eicke-Kanani, L., Fabritz, O., & Wallis, T. S. A. (2025) · Journal of Vision · doi:10.1167/jov.25.8.12
“…(2011) model and extensions by Broderick et al. (2023), the texture tiling model (Balas, Nakano, & Rosenholtz, 2009), and other image quality metrics such as SSIM (Wang, Simoncelli, & Bovik, 2003; Wang et al., 2004), LPIPS (Zhang et al., 2018), and DISTS (Ding, Ma, Wang, & Simoncelli, 2020).”
Unsupervised Insulator Defect Detection Method Based on Masked Autoencoder
Song, Y., & Xiong, W. (2025) · Sensors · doi:10.3390/s25144271
“To enhance perceptual sensitivity, we adopt the multi-scale SSIM (MSSIM) formulation [30], which averages SSIM values over progressively coarser image resolutions.”
Transferable Variational Feedback Network for Vendor and Sequence Generalization in Accelerated MRI
Paul, R., Vora, S., Ding, P. L. K., Patel, A. C., Hu, L., Li, B., & Zhou, Y. (2025) · Big Data Mining and Analytics · doi:10.26599/bdma.2025.9020013
Improved Aggregated Contextual Transformations Based on U-Net for Image Inpainting
Li, S., & Yang, F. O. (2025) · Communications in computer and information science · doi:10.1007/978-981-96-6688-1_3
A Face Sketch Similarity Index for The Evaluation of The Stylized Videos
Zhu, Z., Shu, Z., He, B., Wang, H., Wu, C., & Yang, D. (2025) · International Conference on Computer Vision, Image and Deep Learning (CVIDL) · doi:10.1109/cvidl65390.2025.11085996
“Though multi-scale (MS-SSIM [18]) and gradient-weighted variants (GW-SSIM [10]) enhance perceptual alignment, their computational cost and neglect of semantic abstraction limit applicability.”
Image generation evaluation: a comprehensive survey of human and automatic evaluations
Liu, Q., Yang, S., Li, Z., Hou, L., Meng, C., Zhang, Y., & Sun, L. (2025) · Frontiers of Information Technology & Electronic Engineering · doi:10.1631/fitee.2400904 · Cited by 2
Mpmnet: Multi-Information Perception-Enhanced Mamba Unet for Surface Defect DetectionPreprint
Wan, B., Wang, T., & Lv, C. (2025) · SSRN Electronic Journal · doi:10.2139/ssrn.5375563
An image quality assessment algorithm based on ‘global + local’ feature fusion
Yang, Y., Idris, N. B., Wahab, A. W. A., Yu, D., & Liu, C. (2025) · PeerJ Computer Science · doi:10.7717/peerj-cs.3074
“As highlighted by Wang, Simoncelli & Bovik (2003) , these limitations hinder accurate quality prediction.”
Spatial–Temporal Correlation Information-Based Rate Control for Versatile Video Coding
Zhao, Z., He, X., Xiong, S., et al. (2025) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2025.3595555
“Furthermore, to determine the appropriateness of the scheme for human visual perception, BD-rate calculations based on the Multiscale Structural Similarity Index (MS-SSIM) [71] and the Video Multimethod Assessment Fusion (VMAF) [72] are also applied.”
Lightweight Image Compression Model via Knowledge Distillation
Xu, W., Wang, T., Wang, S., et al. (2025) · Chinese Control and Decision Conference (CCDC) · doi:10.1109/ccdc65474.2025.11090700
“…deep learning [3], some learned image compression methods have outperformed the Versatile Video Coding (VVC) standard [4], which is currently the best traditional image and video coding standard, in terms of both Peak Signal-to-Noise Ratio (PSNR) and Multi-Scale Structural Similarity (MS-SSIM) [5].”
Image Similarity Measures for Autocorrelation Maps Comparison
Přibil, J., & Švehlíková, J. (2025) · International Conference on Measurement (MEASUREMENT) · doi:10.23919/measurement66999.2025.11078643
“Multi-scale structural similarity index multissim [6] incorporates image details at different resolutions and combines the measurement at different scales where c j ( x , y ) and s j ( x , y ) are the contrast comparison and the structure comparison at the j -th scale and l M ( x , y ) is the…”
Predicting Important Photons for Energy-Efficient Single-Photon Videography
Gupta, S., Sundar, V., Koerner, L. J., Bruschini, C. E., Edoardo Charbon, A., & Gupta, M. (2025) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2025.3598767 · Cited by 1
“We compare motion metering on salience-guided sub-sampling to motion metering on uniformly sampled measurements (same sub-sampling factors for all pixels) using the Multi-Scale Structural Similarity (MS-SSIM [61], [62]) and the Feature Similarity (FSIM [63]) metrics.”
Locally Orderless Images for Optimization in Differentiable Rendering
Mehta, I., Chandraker, M. K., & Ramamoorthi, R. (2025) · 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/cvpr52734.2025.00541
“Our experiments are designed for three goals: First, we compare our method against other multi-scale approaches like Gaussian Pyramids [1] and MS-SSIM [38] (§ 4.1).”
InfluentialSketch-Guided Stylized Landscape Cinemagraph SynthesisPreprint
Jin, H., Chang, H., Xie, X., et al. (2025) · SSRN Electronic Journal · doi:10.2139/ssrn.5389750
Self-citationSemantics-Guided Generative Image Compression
Wu, C., Choi, H., & Bajić, I. V. (2025) · Proceedings - International Conference on Image Processing · doi:10.1109/icip55913.2025.11084711
“Pixel-level similarity metrics: Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index Metric (SSIM) [32], and Multi-Scale SSIM (MS-SSIM) [33], which measure pixel-level fidelity relative to the original image; 2.”
X265-PVMAF: A Real-Time Perceptual Video Quality Metric for HEVC Video Encoding
Decker, A. D., Sivashanmugam, S., Cock, J. D., et al. (2025) · Proceedings - International Conference on Image Processing · doi:10.1109/icip55913.2025.11084726
A Style Transfer-Based Fast Image Quality Assessment Method for Image Sensors
Xian, W., Chen, B., Yan, J., Wei, X., Guo, K., Fang, B., & Zhou, M. (2025) · Sensors · doi:10.3390/s25165121 · Cited by 1
“The influence of SSIM led to numerous extensions, including the multiscale SSIM (MS-SSIM) [26], which incorporates multiscale processing to better mimic the HVS, and the 3-SSIM [27], which assigns greater weight to edge regions.”
Liao, C., Todo, Y., Zhang, J., & Tang, Z. (2025) · International Journal of Imaging Systems and Technology · doi:10.1002/ima.70185
“Traditional image similarity metrics such as structural similarity index (SSIM) or peak signal- to-noise ratio (PSNR) are not suitable for this situation as they primarily assess pixel-level similarity and do not capture the complex features inherent in medical images [47–49].”
Spectral Constrained Generative Adversarial Network for Hyperspectral Compression
Guo, Y., Li, W., Peng, Q., & Tu, L.-F. (2025) · IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing · doi:10.1109/jstars.2025.3599996 · Cited by 2
Robustness Verification of Classification Through Image Perturbations
Jha, H. O. K., Singh, A., Pareek, A., Dave, M., Chaudhuri, D., & Singh, M. P. (2025) · International Conference on Range Technology (ICORT) · doi:10.1109/icort64008.2025.11115452
“Wang et al. [9] demonstrated SSIM's effectiveness in capturing structural information changes, while MSE provides a straightforward measure of pixel-level differences.”
Bourbia, S., Karine, A., Chetouani, A., et al. (2025) · Intelligent Systems with Applications · doi:10.1016/j.iswa.2025.200570
Toward Learned Image Compression for Multiple Semantic Analysis Tasks
Tang, Z., Yi, X., & Wang, H. (2025) · IEEE Transactions on Broadcasting · doi:10.1109/tbc.2025.3597154
“First, these DNN-based image compression methods using mean squared error (MSE) or MS-SSIM [10] as object quality metrics [11], [12] focus on optimizing signal fidelity without fully considering the preservation of semantic, which is crucial to machine vision analysis.”
Rate-Distortion-Complexity Optimized Framework for Multi-Model Image Compression
Hang, X., Ge, Z., Fan, H., Jia, C., Ma, S., & Gao, W. (2025) · IEEE Transactions on Image Processing · doi:10.1109/tip.2025.3598916 · Cited by 1
Lossy Compression of Noisy Medical Images
Лукин, В. В., Krylova, O., Kryvenko, S., & Kryvenko, S. (2024) · International Conference on Dependable Systems, Services and Technologies (DESSERT) · doi:10.1109/dessert65323.2024.11122253
“…signal-to-noise ratio (PSNR), visual quality metrics PSNR-HVS-M (https://ponomarenko.info/psnrhvsm.htm, HVS relates to human vision system and M – to masking) and MS-SSIM (multiscale structure similarity [32]) are applied since visual quality of compressed medical images is of prime importance.”
Gujarathi, A., Oza, P., & Bera, A. (2025) · IEEE Access · doi:10.1109/access.2025.3603088
“Subsequently, Gaussian noise is introduced to complement GAN image and boost robustness of the encrypted image metadata [55].”
Finch-LIC: Learned Image Compression with Gated Multihead Linear AttentionPreprint
Yi, F., Gong, Z., & Zeng, H. M. (2025) · SSRN Electronic Journal · doi:10.2139/ssrn.5413219
Acceleration of Optimization-based Structured Sparse Time-Frequency Analysis by ADMM
Arai, K., & Yatabe, K. (2025) · International Conference on Sampling Theory and Applications · doi:10.1109/sampta64769.2025.11133562
“As an evaluation metric, we adopted the multi-scale structural similarity (MS-SSIM) [31] that quantifies image similarity based on luminance, constrast, and structure as it reflects similarity in terms of structural properties.”
Visual Perception‐Based Quality Assessment for Stitched Panoramic Images
Fan, Y., Chen, C., Hu, X., Li, Y., Yu, H., Zhang, R., & Han, L. (2025) · IET Image Processing · doi:10.1049/ipr2.70198 · Cited by 1
Liu, X., Chen, J., Chen, B., Liu, Z., An, B., Xia, S., & Wang, Z. (2025) · IEEE Transactions on Multimedia · doi:10.1109/tmm.2025.3604982 · Cited by 1
“These advancements have led to significant progress, outperforming many traditional image codecs such as JPEG [5] and BPG [6] across common metrics like PSNR and MS-SSIM [7].”
BackgroundSelf-citationSelf-Attention Based Motion Error Correction for Neural Video CompressionPreprint
Chen, J., Zhao, Q., Zeng, H., et al. (2025) · SSRN Electronic Journal · doi:10.2139/ssrn.5437324
Black-Box Attacks on Interpretable Deepfake Detectors
Kanwal, P., Agrawal, I., Dhanalakshmi, K. M., Hamsini, S., Bhavana, S., & Honnvalli, P. B. (2025) · International Conference for Emerging Technology (INCET) · doi:10.1109/incet64471.2025.11140391
Amadin, V., Chung, M., Kim, Y., et al. (2025) · SSRN Electronic Journal · doi:10.2139/ssrn.5449640
Multimodal Medical Image Fusion With UNet ‐Based Multi‐Scale Transformer Networks
Fu, Q., Zhou, J., Du, J., Lin, K., Zhong, B., Tang, H., & Chen, Y.-T. (2025) · International Journal of Imaging Systems and Technology · doi:10.1002/ima.70193
“…performance: nonlinear correlation information entropy ( Q NCIE ) [33], gradient-based metric ( Q G ) [34], phase congruence-based image feature metric ( Q P ) [35], Chen-Varshney metric ( Q CV ) [36], Chen-Blum metric ( Q CB ) [37], and multi-scale structural similarity index (MS-SSIM) [38].”
Benchmarking Learnable Mesh and Texture Representations for Immersive Digital Twins
Müller, L. P., Bätz, M., Berg, A., et al. (2025) · IEEE International Conference on Multimedia and Expo workshops · doi:10.1109/icmew68306.2025.11152025
“We provide PSNR, MS-SSIM [34], and LPIPS [35] metrics for comparison.”
Transformer-Based Task-Oriented Joint Source-Channel Coding for Edge AI Agent
Du, J. J., She, C., Geng, Z., et al. (2025) · IEEE/CIC International Conference on Communications in China (ICCC) · doi:10.1109/icccworkshops67136.2025.11148093
“…quality of reconstructed images using three standard metrics: • PSNR (Peak Signal-to-Noise Ratio) • MS-SSIM (Multi-Scale Structural Similarity) [23] • LPIPS (Learned Perceptual Image Patch Similarity) [24] These metrics help quantify the visual degradation introduced by different JSCC methods…”
Event-Based Tracking and Imaging of Randomly Moving Objects in Dense Dynamical Scattering Media
Zhang, N., Shea, T. M., & Nurmikko, A. V. (2025) · IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw67362.2025.00507
“To achieve accurate and high-quality reconstructions, globally and pixel-wise, a hybrid loss function is employed, comprising three components: (a) Binary Cross-Entropy (BCE) Loss[43] evaluates pixel-level accuracy by comparing each pixel in the reconstruction with the corresponding ground truth value.”
Integrating Homomorphic Encryption and Synthetic Data in FL for Privacy and Learning Quality
Wang, Y., Chiasserini, C. F., & Schiller, E. M. (2025) · IEEE International Symposium on Local and Metropolitan Area Networks (LANMAN) · doi:10.1109/lanman66415.2025.11154574
Tsukijima, M., Teramoto, A., Kojima, A., et al. (2025) · Radiological Physics and Technology · doi:10.1007/s12194-025-00949-4
HOPE: Enhanced Position Image Priors via High-Order Implicit Representations
Chen, Y., Wu, R., Hou, J., et al. (2025) · IEEE Transactions on Image Processing · doi:10.1109/tip.2025.3607582
“We evaluate performance using Peak Signal-to-Noise Ratio (PSNR) and Structural Similarity Index Measure (SSIM) [48].”
Texture Optimization with Preserved ResolutionPreprint
Chen, W., Bai, W., Ren, Y., et al. (2025) · SSRN Electronic Journal · doi:10.2139/ssrn.5503640
A Framework With Multi-Scale Hybrid Mamba Voxel Flow for Video Prediction
Xu, M., Fu, B., Liu, D., Deng, W., Yi, W., Wan, Y., Hua, W., & Song, W. (2025) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2025.3612980
“To evaluate model performance, we use multi-scale structural similarity (MS-SSIM) [40] and learned perceptual image patch similarity (LPIPS) [41].”
Textured Mesh Quality Assessment Using Geometry and Color Field Similarity
Yang, K., Yang, Q., Xu, Y., et al. (2025) · IEEE Transactions on Visualization and Computer Graphics · doi:10.1109/tvcg.2025.3612942
“For video-based metrics, PSNR, SSIM [31], MS-SSIM [34], 3-SSIM [35], VQM [36], and VMAF [37] are calculated using the MSU Video Quality Measurement Tool [38] based on the rendered videos used in subjective experiments.”
Narshihbhai, J. K., & Bhat, A. (2025) · 2025 International Conference on Computing Technologies & Data Communication (ICCTDC) · doi:10.1109/icctdc64446.2025.11158981
“• L msssim : Multi-Scale SSIM loss [13] calculated as 1 − MS-SSIM ( ˆ S, S norm ) .”
DALL-E Brain: Generating 2D T1-w, T2-w, and FLAIR Brain MRI Images from Textual Prompts
El-Allaly, S., Grenier, T., & Revol-Muller, C. (2025) · European Signal Processing Conference (EUSIPCO) · doi:10.23919/eusipco63237.2025.11226818
“To objectively assess reconstruction fidelity, we use the Multi-Scale Structural Similarity Index Measure (MS-SSIM) [12], which evaluates image similarity across multiple resolutions, incorporating luminance, contrast, and structural components.”
Image quality assessment: the integral CTS index for holistic similarity evaluation of images (ISII)
Ivković, R., & Milivojević, Z. (2025) · Multimedia Tools and Applications · doi:10.1007/s11042-025-21120-6 · Cited by 1
M-JSCC: An Asymmetric Semantic Communication Architecture for 6G Intelligent Networks
Ren, P., Wang, J., Wang, Z., Hou, X., Zhang, X., & Jiang, C. (2025) · IEEE Vehicular Technology Conference · doi:10.1109/vtc2025-spring65109.2025.11174786
“For image reconstruction, we use the peak signal-to-noise ratio (PSNR) and multi-scale structural similarity (MS-SSIM) as metrics [13].”
Self-citationDiffusion-based SDRTV-to-HDRTV Reconstruction with Artifact SuppressionPreprint
Li, X., Weimin, Z., Huiyuan, F., & Ma, H. (2025) · SSRN Electronic Journal · doi:10.2139/ssrn.5549060 · Cited by 1
Document Rectification via Hierarchical Attention-Guided Control Point Regression
Ni, H., & Han, J. (2025) · International Conference on Computer Information Science and Application Technology (CISAT) · doi:10.1109/cisat66811.2025.11181731
“…we employ four quantitative metrics that assess both geometric rectification accuracy and text preservation quality: • Multi-Scale Structural Similarity (MS-SSIM) [10]measures the global structural similarity between the rectified image and the ground-truth document at multiple spatial scales.”
Ailuro, S. M., Nedorubova, A., Grigoryev, T., Burnaev, E. V., & Vanovskiy, V. (2025) · Scientific Reports · doi:10.1038/s41598-025-16572-7
“In addition to this, we utilize the Structural Similarity Index (SSIM)[16] and its extension, the Multi-Scale Structural Similarity Index (MS-SSIM)[17], to assess the perceived quality of digital images and videos.”
Q-Space Guided Multi-Modal Translation Network for Diffusion-Weighted Image Synthesis
Zhu, P., Fu, Y., Chen, N., & Qiu, A. (2025) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2025.3618683 · Cited by 1
“In alignment with previous experiments, we evaluate the synthesized DWI using several image quality metrics: RMSE, MS-SSIM [54], PSNR, and LPIPS [55].”
NVENC Codecs Quality Comparison for Real-Life UAV 4k Video Streams
Vasile, C.-E., & Bîră, C. (2025) · IEEE International Black Sea Conference on Communications and Networking · doi:10.1109/blackseacom65655.2025.11193943
“SSIM VISUAL QUALITY METRIC Paper [8] proposes a multi-scale structural similarity metric for image quality assessment by evaluating how similar a distorted or compressed image is to a reference (original) image, aiming to reflect perceived visual quality more accurately than traditional metrics…”
Performance comparison of different Retinex methods on enhancement of low light cave images
Yazlık, B., & Yazlık, E. N. (2025) · Ömer Halisdemir Üniversitesi Mühendislik Bilimleri Dergisi · doi:10.28948/ngumuh.1776861
High Resolution SAR to Multispectral Image Translation for Urban BuildingsPreprint
Han, Q., Liu, Y., Hong, Y., Zhou, D., & Xing, M. (2025) · SSRN Electronic Journal · doi:10.2139/ssrn.5625503
Zhang, Y., Wang, Z., Kong, X., Luo, Y., Zhang, X., Chen, M., Chen, S., Shen, Y., Lin, H., & Chen, X. (2025) · IEEE International Ultrasonics Symposium · doi:10.1109/ius62464.2025.11201762
“We assess PD-to-ULM image translation performance with structural metrics (MS-SSIM [12], PSNR [13], DSC) and perceptual quality metrics (three LPIPS scores from AlexNet, VGG-16, SqueezeNet).”
Self-citationCellular Aggregation Graph Convolutional Network for Point Cloud Quality Assessment
Xiong, J., Jiang, L., Hu, Q., et al. (2025) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2025.3624798
“Previous studies in IQA [19], [20], [21], [22] have established that the human visual system (HVS) exhibits remarkable adaptability to extract hierarchical structural information.”
OpenMVC: An Open-Source Library for Learning-based Multi-view Compression
Zheng, H., & Gao, W. (2025) · ACM International Conference on Multimedia (ACM MM) · doi:10.1145/3746027.3756882
Evaluating Intra-Class Diversity Using MS-SSIM: Insights from the NEU DatasetPreprint
Tien, H. V., & Le, C. T. D. (2025) · SSRN Electronic Journal · doi:10.2139/ssrn.5679810
A Comprehensive Dataset for Image Segmentation in Custom Manufacturing Environments
Bell, M., & Vitali, R. V. (2025) · Scientific Data · doi:10.1038/s41597-025-06007-3 · Cited by 1
Weak Semantic-Guided Entropy Model for Image Compression
Ding, Y., & Wei, J. (2025) · IEEE International Conference on Multimedia and Expo (ICME) · doi:10.1109/icme59968.2025.11208925
“To optimize the proposed method, the loss function is defined as: where D ( x, ˆ x ) represents the distortion between the input image x and the reconstructed image ˆ x , measured by either MSE or MS-SSIM [30].”
InfluentialSpectrum-Adaptive Distribution of 2D Gaussians for Image Representation and Compression
Wan, Z., Zhao, J., Ding, Y., Zhang, L., Satō, H., & Ogawa, T. (2025) · IEEE International Conference on Multimedia and Expo (ICME) · doi:10.1109/icme59968.2025.11209121
“Evaluation metrics include PSNR and MS-SSIM [22] for reconstruction quality, bits per pixel (bpp) for compression efficiency, and encoding/decoding speed (FPS) for computational performance.”
IMPACT OF LOSS FUNCTION ON SYNTHETIC BREAST ULTRASOUND IMAGE GENERATION
Ryspayeva, M., & Salykova, O. (2025) · Scientific Journal of Astana IT University · doi:10.37943/24mmik3887
DCU: A New Dynamic Codebook Update Improvement on the Vector Quantization Model
Xu, Y., Shen, L., Cui, C., et al. (2025) · IEEE International Conference on High Performance Computing and Communications (HPCC) · doi:10.1109/hpcc67675.2025.00203
“In terms of evaluation metrics, we adopt widely-used image quality assessment metrics, including the average Peak Signal-to-Noise Ratio (PSNR), Learned Perceptual Image Patch Similarity (LPIPS) [13], and Multi-Scale Mean Structural Similarity Index Measure (MS-SSIM) [14].”
An Adaptive Measurement Method Combining MSE and SSIM for Image Transmission
Cui, H., Qiu, S., Xu, J., & Xu, Y. (2025) · IEEE International Conference on High Performance Computing and Communications; IEEE International Conference on Smart City; IEEE International Conference on Data Science and Systems (HPCC/SmartCity/DSS) · doi:10.1109/hpcc67675.2025.00204
Accelerating Learned Video Compression via Low-Resolution Representation Learning
Qiu, Z., He, Z., & Jin, Z. Y. (2025) · Lecture notes in computer science · doi:10.1007/978-981-95-3398-5_36
Generating adversarial examples based on global frequency domain to attack object detectorsPreprint
Nguyen, D., & Hùng, P. N. (2025) · SSRN Electronic Journal · doi:10.2139/ssrn.5699161
An Improved U-Net Deep Learing Method to Reduce MRI Gadolinium-Based Contrast Agents
Shukur, N. J., Ibrahim, M., El‐Zaart, A., et al. (2025) · International Conference on Advances in Biomedical Engineering (ICABME) · doi:10.1109/icabme66883.2025.11211790
Dual-Path Divide-and-Conquer Neural Network for Image Deblurring
Shin, J., & Lee, H.-S. (2025) · The Journal of Korean Institute of Information Technology · doi:10.14801/jkiit.2025.23.10.185
ν GAN: A Deep Learning Emulator for Cosmic Web Simulations with Massive Neutrinos
Kaushal, N., Giusarma, E., & Reyes, M. (2026) · The Astrophysical Journal · doi:10.3847/1538-4357/ae3de4 · Cited by 1
Learning-Based Compressed-Domain Texture Recognition
Ji, R., & Karam, L. J. (2025) · IEEE International Conference on Multimedia Information Processing and Retrieval (MIPR) · doi:10.1109/mipr67560.2025.00074
“In this paper, the variational image compression with a hy-perprior [17] and with a MS-SSIM [24] reconstruction quality metric in the loss function (HyperMS-SSIM) is employed.”
Direction-Oriented Edge Reconstruction Network for Camouflaged Object Detection
Yang, X., Zhang, Q., & Zhou, Q. (2025) · doi:10.1109/ijcnn64981.2025.11227430
“BASNet [14] designed a hybrid loss combining binary cross-entropy loss [19], structural similarity loss [20], and IoU loss [21] to implicitly guide the network to focus more on the boundary information of the object.”
Controllable-Quality Neural Video Representation via Inter-frame Motion and Residual Mapping
Wan, Z., & Zhang, D. (2025) · doi:10.1109/ijcnn64981.2025.11227918
“For evaluation, we adopted PSNR and MS-SSIM [30] as our performance metrics.”
ConvGAN: A ConvNeXt-Enhanced CycleGAN for Chinese Calligraphy and Painting Fusion
Peng, Y., Xiong, M., & Zhao, Q. (2025) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn64981.2025.11228607
A novel lightweight network VMobileNet for visual detection of surface roughness in turning
Liu, X.-C., Li, Y., Shi, L., et al. (2025) · Measurement science and technology · doi:10.1088/1361-6501/ae2044
TopoRN: A topology-aware reconstruction network for infrared image super-resolution
Dan, J., Jin, T., Chi, H., Luo, Z., Cao, K., Xinjing, Y., & Yang, X. (2025) · Applied Intelligence · doi:10.1007/s10489-025-06999-w · Cited by 1
“However, because MSE relies on pixel-wise image dis - parities, its usefulness as a metric to capture perceptually significant differences, such as object edges and high-fre - quency details, is severely restricted [ 1, 69].”
Explainable and likelihood aware AI framework for MRI-based pixel-level bladder tumour prediction
Khan, M., de Groot, A. G., Cornel, E. B., van der Heijden, A. G., & Siepel, F. J. (2025) · Scientific Reports · doi:10.1038/s41598-025-24647-8 · Cited by 1
“SSIM loss 41 ensures structural alignment accuracy, complemented by a gradient loss 42 to enforce smooth deformations along with an uncertainty loss.”
NajVR: low-latency plug-and-play remote rendering for interactive virtual reality applications
Santos-Paz, J. A., Aranda, J. J. G., Granero, M. A., et al. (2025) · Computing · doi:10.1007/s00607-025-01589-7
“Additionally, to better compared the quality of the encoded videos, MS-SSIM [ 24] was also used (see Fig.”
Human-Aware Inpainting for 3D Gaussian Splatting in Dynamic Scenes
Liao, P.-R., & Kawai, N. (2025) · IEEE International Symposium on Mixed and Augmented Reality Adjunct (ISMAR-Adjunct) · doi:10.1109/ismar-adjunct68609.2025.00197
“The loss function follows the original 3DGS formulation, combining point-wise L1 loss and SSIM [11]: where Here, R i refers to the rendered images, while H and W denote the height and width of the input image I , respectively.”
Anomaly Detection in Anterior Eye Segment Using Self-supervised Siamese Autoencoders
Baumstimler, P., Gagné, S., Champagne, M., Gagnon, J.-M., Playout, C., & Séoud, L. (2025) · Lecture notes in computer science · doi:10.1007/978-3-032-10351-2_5
An Infrared and Visible Image Fusion Strategy Integrating Global Multi-Scale Information
Li, J., Li, X., & Chen, S. (2025) · IEEE International Geoscience and Remote Sensing Symposium (IGARSS) · doi:10.1109/igarss55030.2025.11243544
“On the other hand, MS-SSIM[13], a multi-scale analysis metric for image structure, is more effective in capturing the comprehensive structural features of images.”
Lisowski, C. M., King, G. M., & Kosztin, I. (2025) · PLoS Computational Biology · doi:10.1371/journal.pcbi.1013277
“The quality of LAFM images relative to the benchmark (i.e., the LAFM images AP–2–2 and AC–2–2, respectively) was evaluated using the Structural Similarity Index Measure (SSIM) [30], which offers several advantages over simpler metrics such as mean squared error [31].”
Lightweight Deep Learning-Based Aortic Valve Segmentation on RGB Images
Fromme, D., Streckert, T., Warnakulasuriya, D., Schuldt, D., Bogatzki, M., Gomez, F. J. C., & Thiem, J. (2025) · Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC) · doi:10.1109/embc58623.2025.11253048
“Our objective is to evaluate the encoder’s feature extraction capability by measuring reconstruction quality using the following metrics: Mean Squared Error (MSE), Mean Absolute Error (MAE), SSIM, Multi-Scale SSIM (MS-SSIM)[16] and KID [17].”
Class-N-Diff: Classification-Induced Diffusion Model Can Make Fair Skin Cancer Diagnosis
Munia, N., & Imran, A.-A.-Z. (2025) · Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC) · doi:10.1109/embc58623.2025.11252934
“Classification Performance: We train the classification model independently and compare its performance with Authorized licensed use limited to the terms of the applicable license agreement with IEEE.”
3D Conditional Image Synthesis of Left Atrial LGE MRI from Composite Semantic Masks
Al-Sanaani, Y., Thornhill, R. E., & Rajan, S. (2025) · IEEE International Conference on Imaging Systems and Techniques (IST) · doi:10.1109/ist66504.2025.11268401
Sparse SVBRDF Acquisition via Importance-Aware Illumination Multiplexing
Zhang, L., Wang, Z., Wang, L., et al. (2025) · ACM Transactions on Graphics · doi:10.1145/3763324
“Instead, we use Multi-Scale Structural Similarity Index Measure (MS-SSIM) [Wang et al. 2003] to measure the structure difference between the sum of all predicted patterns { 𝐼 } and 𝐿 𝑝𝑑𝑓 .”
Self-citationZS-KAN: Zero-Shot Image Denoising With Lightweight Kolmogorov--Arnold Networks
Wang, J., & Wang, G. (2025) · IEEE Transactions on Biomedical Engineering · doi:10.1109/tbme.2025.3640551 · Cited by 2
“Theevaluationmetricsarepeaksignal-to-noiseratio (PSNR), structural similarity (SSIM) [32], and multi-scale structural similarity (MS-SSIM) [33].”
Optimizing Image Compression: Perspectives on SVD-Based Restoration
Chen, J., Chen, P., Wang, S., et al. (2025) · European Signal Processing Conference (EUSIPCO) · doi:10.23919/eusipco63237.2025.11226105
“To evaluate the R-D performance, we measure the rate in bits per pixel (bpp) and assess the quality of the decompressed images using PSNR and MS-SSIM [20].”
Joint Optimization for Image Compression and Deblurring via Blur-Aware Guidance
Wan, Z., Zhao, J., Ding, Y., Guan, J., Zhang, L., & Ogawa, T. (2025) · ACM International Conference on Multimedia in Asia (MMAsia) · doi:10.1145/3743093.3770957
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.