The DOI matches its Crossref record; 56 of the 56 references it cites were found in a registry or an index; none of the 25 sources its claims cite is checked for support yet.
No-Reference Image Quality Assessment in the Spatial Domain
Abstract
We propose a natural scene statistic-based distortion-generic blind/no-reference (NR) image quality assessment (IQA) model that operates in the spatial domain. The new model, dubbed blind/referenceless image spatial quality evaluator (BRISQUE) does not compute distortion-specific features, such as ringing, blur, or blocking, but instead uses scene statistics of locally normalized luminance coefficients to quantify possible losses of "naturalness" in the image due to the presence of distortions, thereby leading to a holistic measure of quality. The underlying features used derive from the empirical distribution of locally normalized luminances and products of locally normalized luminances under a spatial natural scene statistic model. No transformation to another coordinate frame (DCT, wavelet, etc.) is required, distinguishing it from prior NR IQA approaches. Despite its simplicity, we are able to show that BRISQUE is statistically better than the full-reference peak signal-to-noise ratio and the structural similarity index, and is highly competitive with respect to all present-day distortion-generic NR IQA algorithms. BRISQUE has very low computational complexity, making it well suited for real time applications. BRISQUE features may be used for distortion-identification as well. To illustrate a new practical application of BRISQUE, we describe how a nonblind image denoising algorithm can be augmented with BRISQUE in order to perform blind image denoising. Results show that BRISQUE augmentation leads to performance improvements over state-of-the-art methods. A software release of BRISQUE is available online: http://live.ece.utexas.edu/research/quality/BRISQUE_release.zip for public use and evaluation.
Overview
AI snapshot
Written from the paper’s full text.
TL;DR
BRISQUE is a no-reference, distortion-generic image quality model that operates in the spatial domain by fitting natural scene statistics of locally normalized luminance coefficients and their pairwise products, then mapping features to quality with SVR. It is statistically better than PSNR and SSIM, competitive with NR methods, efficient, and useful for blind denoising.
Key findings
- On the LIVE IQA database, BRISQUE achieved median SROCC 0.9395 and LCC 0.9424 across 1000 train-test splits; it was statistically better than full-reference PSNR and SSIM, highly competitive with NR algorithms, and slightly inferior to FR MS-SSIM.
- BRISQUE features also supported distortion classification, with median accuracy 88.6% overall: WN 100.0%, Blur 96.7%, JPEG 88.9%, JP2K 82.9%, and FF 83.3%; FF and JP2K were most confused.
- Computational complexity was low: for a 512 × 768 image on a 1.8-GHz single-core PC with 2 GB RAM, BRISQUE took 1 s, versus DIIVINE 149 s, BLIINDS-II 70 s, and PSNR 0.05 s.
- BRISQUE-augmented BM3D blind denoising was statistically superior to reference BM3D at the 95% confidence level for all simulated noise variances except 0.0316; on TID2008, BRISQUE’s overall SROCC was 0.896, compared with PSNR 0.870 and SSIM 0.902.
Methods
- The model computes mean subtracted contrast normalized (MSCN) coefficients via local mean subtraction and divisive normalization, fits a GGD to MSCN coefficients, and fits AGGD models to pairwise products along horizontal, vertical, main-diagonal, and secondary-diagonal orientations. It extracts 36 features (18 per scale at two scales) and maps them to quality using an SVM regressor with an RBF kernel.
- Evaluation used the LIVE IQA database (29 reference images, 779 distorted images, five distortion categories), with random 80% training / 20% testing splits, no content overlap, 1000 repetitions, and median SROCC/LCC; statistical significance was assessed with one-sided t-tests.
- Database independence was tested by training on all of LIVE and applying BRISQUE to TID2008; blind denoising used 300 Berkeley images, 10 Gaussian noise levels (3000 noisy images), 1000 training and 2000 testing images, with BRISQUE features predicting the BM3D input noise parameter.
Limitations
- Stated: BRISQUE remains slightly inferior to FR MS-SSIM, indicating room for improvement; the two-stage version showed a slight dip attributed to imperfect distortion identification; TID2008 testing used only 24 natural images and only distortions BRISQUE was trained for (JPEG, JP2K, WN, blur; FF absent); performance decreases when the window size becomes fairly large.
- Evident from the text: BRISQUE requires a training/calibration procedure with subjective quality scores, and its evaluation and denoising application depend on known distortion or noise conditions used in training.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
21 Own result · 14 Not checked yet · 14 No basis given · 12 Own interpretation · 6 Sources not found · 4 Own method
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
“It can be visualized that different distortions occupy different parts of the space.”
plaincorrelational- 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
“As shown in Fig. 5(a), pristine and distorted images occupy different regions in this parameter space.”
plaincorrelational- 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
“Despite its simplicity, we are able to show that BRISQUE is statistically better than the full-reference peak signal-tonoise ratio and the structural similarity index, and is highly competitive with respect to all present-day distortion-generic NR IQA algorithms.”
boosted- 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
“We observed that increasing the number of scales beyond 2 did not contribute to performance much.”
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.
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
“Results show that BRISQUE augmentation leads to performance improvements over state-of-the-art methods.”
plaindirect- 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
“Despite its simplicity, we are able to show that BRISQUE is statistically better than the full-reference peak signal-tonoise ratio and the structural similarity index, and is highly competitive with respect to all present-day distortion-generic NR IQA algorithms.”
boosted- 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
“We have found that a generalized Gaussian distribution (GGD) can be used to effectively capture a broader spectrum of distorted image statistics, which often exhibit changes in the tail behaviour (i.e. kurtosis) of the empirical coefficient distributions [32]”
What the sources say
- Not checked yetcites [32] [32] K. Sharifi and A. Leon-Garcia, 'Estimation of shape parameter for generalized Gaussian distributions in subband decompositions of video,' IEEE Trans. Circuits Syst. Video Technol. , vol. 5, no. 1, pp. 52-56, Feb. 1995. - source
plaincorrelational- 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.
- Rests on an assumption: MSCN coefficient distributions are symmetric.
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
“We note that BRISQUE remains slightly inferior to the FR MS-SSIM, indicating that there may still be some room for improvement in performance.”
hedged- 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
“The performance of the algorithm drastically improves, if a (systematically) different parameter selected based on perceptual quality is fed as input to the algorithm.”
plaindirect- 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
“Further, we also showed that BRISQUE is computationally efficient and that its efficiency is superior to other distortion-generic approaches to NR IQA, thus making BRISQUE an attractive option for practical applications like image denoising.”
boosted- 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.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“In this section, we describe one such approach, where the BRISQUE features are used to transform a nonblind image denoising algorithm into a blind image denoising algorithm.”
- Also stated in: § VI. CONCLUSION · p. 13 · boosted
“We demonstrated this application by augmenting non-blind image denoising algorithms using the BRISQUE features to produce blind image denoising algorithms.”
boosteddirect- 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.
- Rests on an assumption: Existing blind image denoising algorithms typically address parameter estimation in an ad-hoc fashion without regard to natural scene statistics.
- Also stated in: § VI. CONCLUSION · p. 13 · boosted
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
“It should be clear that BRISQUE performs well in terms of correlation with human perception of quality and that the performance does not depend on the database.”
boostedcorrelational- 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
“We detailed the algorithm and the statistical features extracted, and demonstrated how each of these features correlate with human perception.”
boostedcorrelational- 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
“It is clear that BRISQUE augmented BM3D produces much higher quality images than the baseline BM3D.”
- Also stated in: § V. APPLICATION TO BLIND IMAGE DENOISING · p. 12 · boosted
“Given a new (unseen) test noisy image, the BRISQUE augmented BM3D approach predicts the accurate input to BM3D and denoises the image with (as we shall soon see) much higher visual quality than the baseline.”
boosteddirect- 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: § V. APPLICATION TO BLIND IMAGE DENOISING · p. 12 · boosted
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“that demonstrates better ability to predict human judgments of quality than other popular FR and NR IQA models; that is highly efficient; and that is useful for perceptually optimizing image processing algorithms such as denoising.”
boosted- 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: unclear — It is unclear why the paper says it holds.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“We see from Fig. 10 that FF and JP2K are most confused with each other which is not surprising, since FF distortion is a combination of JP2K followed by packet-loss errors.”
plaincorrelational- 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.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“In Table VII, we tabulate the median SROCC value across 1000 trials for the two-stage realization of BRISQUE. We also list the performances of BRISQUE for comparison purposes. The slight dip in the performance can be attributed to imperfect distortion identification in the first stage of the two-stage framework.”
plaindirect- 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
“However, our experiments suggest that when the algorithm is fed the true accurate noise-variance, the performance of the denoiser is sub-par.”
hedgeddirect- 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.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“Here, we demonstrate a systematic perception-based parameter estimation approach that results in better denoising performance.”
boosteddirect- 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: unclear — It is unclear why the paper says it holds.
- Rests on an assumption: Existing blind image denoising algorithms typically address parameter estimation in an ad-hoc fashion without regard to natural scene statistics.
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
“WN and Blur are generally not confused with other distortions.”
plain- 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
“We show that BRISQUE performs well on independent databases, analyze its complexity and compare it with other NR IQA approaches.”
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.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“Once trained, the automatic parameter prediction approach is capable of predicting the level of input noise to BM3D, so that the output denoised image has the highest visual quality.”
plaindirect- 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
“In order to demonstrate that BRISQUE features can also be used for explicit distortion-identification [48], we report the median classification accuracy of the classifier for each of the distortions in the LIVE database, as well as across all distortions in Table VI.”
- Also stated in: § VI. CONCLUSION · p. 13 · boosted
“We demonstrated that BRISQUE performance is independent of database content and BRISQUE features may be used for distortion-identification as well.”
What the sources say
- Not checked yetcites [48] [48] A. K. Moorthy and A. C. Bovik, 'A two-stage framework for blind image quality assessment,' in Proc. IEEE Int. Conf. Image Process. , Sep. 2010, pp. 2481-2484. -
boostedconditional- 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: § VI. CONCLUSION · p. 13 · boosted
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Deviations from the regularity of natural statistics, when quantified appropriately, enable the design of algorithms capable of assessing the perceptual quality of an image without the need for any reference image.”
plaindirect- 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: unclear — It is unclear why the paper says it holds.
In the paper
“Moreover, the assumption of a pristine nature of any reference is questionable, since all images are ostensibly distorted [14].”
What the sources say
- Not checked yetcites [14] [14] A. C. Bovik, 'Perceptual image processing: Seeing the future,' Proc. IEEE , vol. 98, no. 11, pp. 1799-1803, Nov. 2010. - source
plain- Asserted: passes — The authors put it forward as true.
- Checkable: unclear — It is unclear whether it could 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 conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“By quantifying natural image statistics and refraining from an explicit characterization of distortions, our approach to quality assessment is not limited by the type of distortions that afflict the image.”
plaindirect- 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.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“Since the BRISQUE approach requires a training procedure to calibrate the regressor module, we divide the LIVE database into two randomly chosen subsets - 80% training and 20% testing - such that no overlap between train and test content occurs.”
plaindirect- 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
“Under the Gaussian coefficient model, and assuming the MSCN coeffficients are zero mean and unit variance, these products obey the following distribution in the absence of distortion [34]:”
What the sources say
- Not checked yetcites [34] [34] A. H. Nuttall, 'Accurate efficient evaluation of cumulative or exceedance probability distributions directly from characteristic functions,' Naval Underwater Systems Center, New London, CT, Tech. Rep. ADA133703, 1983. - source
plaincorrelational- 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.
- The authors state this as an assumption: The MSCN coefficients are zero mean and unit variance.
None of the sources cited for this claim have been checked yet.
In the 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].”
What the sources say
- Not checked yetcites [13] [13] M. Saad, A. C. Bovik, and C. Charrier, 'Blind image quality assessment: A natural scene statistics approach in the DCT domain,' IEEE Trans. Image Process. , vol. 21, no. 8, pp. 3339-3352, Aug. 2012. - source
- Not checked yetcites [29] [29] Z. Wang, E. P. Simoncelli, and A. C. Bovik, 'Multiscale structural similarity for image quality assessment,' in Proc. Asilomar Conf. Signals, Syst. Comput. , vol. 2. 2003, pp. 1398-1402. -
boosteddirect- 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
“Such approaches to NR IQA are significant since most current approaches are distortion-specific [5]-[11], i.e., they are capable of performing blind IQA only if the distortion that afflicts the image is known beforehand, e.g., blur or noise or compression and so on (see below).”
What the sources say
- Not checked yetcites [5]-[11]
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.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“BRISQUE introduces a new model of the statistics of pair-wise products of neighboring (locally normalized) luminance values.”
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.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“that is useful for perceptually optimizing image processing algorithms such as denoising.”
plaindirect- 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: unclear — It is unclear why the paper says it holds.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Having evaluated BRISQUE on the LIVE IQA database, we now demonstrate that the performance of BRISQUE is not bound by the database on which it is tested.”
- Also stated in: § F. Database Independence · p. 11 · boosted
“It should be clear that BRISQUE performs well in terms of correlation with human perception of quality and that the performance does not depend on the database.”
- Also stated in: § VI. CONCLUSION · p. 13 · boosted
“We demonstrated that BRISQUE performance is independent of database content and BRISQUE features may be used for distortion-identification as well.”
boosted- 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: § F. Database Independence · p. 11 · boosted
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“This suggests that the spatial-domain BRISQUE an ideal candidate for real-time blind assessment of visual quality.”
hedged- 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 conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“This result suggests that to the extent distortions can be trained on, one can replace full reference algorithms such as SSIM with the proposed BRISQUE without any loss of performance.”
hedged- 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.
- It may state two separate claims as one. 23
- Probably not a substantial claim. 18
- The quote only partly states this claim. 1
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“While MSCN coefficients are definitely more homogenous for pristine images, the signs of adjacent coefficients also exhibit a regular structure, which gets disturbed in the presence of distortion.”
plaincorrelational- 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.
It may state two separate claims as one.
In the paper
“JPEG2000 and fast fading have a high degree of overlap as fast fading images in LIVE database are actually multidistorted, first compressed into a bitstream using a JPEG2000 codec, then passed through a Rayleigh fast fading channel to simulate packet loss [27].”
What the sources say
- Not matchedcites [27] [27] H. R. Sheikh, M. F. Sabir, and A. C. Bovik, 'A statistical evaluation of recent full reference image quality assessment algorithms,' IEEE Trans. Image Process. , vol. 15, no. 11, pp. 3440-3451, Nov. 2006. -
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
plaindirect- 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.
It may state two separate claims as one.
- Not matched
In the paper
“The normalized luminance coefficients (explained later) of the natural image closely follow Gaussian-like distribution, as shown in Fig. 1(c) while the same doesnot hold for the empirical distribution of the artificial image shown in Fig. 1(d).”
What the sources say
- Not matchedcites [4] [4] N. Ponomarenko, V. Lukin, A. Zelensky, K. Egiazarian, M. Carli, and F. Battisti, 'TID2008-a database for evaluation of full-reference visual quality assessment metrics,' Adv. Modern Radioelectron. , vol. 10, no. 10, pp. 30-45, 2009. -
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
plaincorrelationalSingle case- 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.
It may state two separate claims as one.
- Not matched
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“BRISQUE has very low computational complexity, making it well suited for real time applications.”
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds. It may state two separate claims as one.
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 Fig. 8, we plot the Spearman's rank ordered correlation coefficient (SROCC) between each of these features and human DMOS from the LIVE IQA database, for each of the distortions in the database JPEG and JPEG2000 compression, additive white Gaussian noise, Gaussian blur and a Rayleigh fast fading channel distortion, to ascertain how well the features correlate with human judgments of quality.”
plaincorrelational- 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.
The quote only partly states this claim.
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
“While we have found that this density function is a good model of the empirical histograms of products of adjacent normalized coefficients, it has only a single parameter, and as such, does not provide a good fit to the empirical histograms of coefficient products (Fig. 2) from distorted images. Further, it is not finite at the origin.”
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.
It may state two separate claims as one.
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
“White noise is very clearly separated from the pristine image set making it one of the easiest to gauge the quality of”
boostedcorrelational- 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.
It may state two separate claims as one.
In the paper
“From Table V we conclude that BRISQUE is highly competitive with all no reference algorithms tested and statistically better than the full reference algorithms PSNR and SSIM.”
- Also stated in: § VI. CONCLUSION · p. 13 · boosted
“We then undertook a thorough evaluation of the BRISQUE index in terms of correlation with human perception and demonstrated that BRISQUE is statistically better than FR PSNR and SSIM as well as highly competitive to all NR algorithms compared with.”
What the sources say
- Not matchedcites [47] [47] D. Sheskin, Handbook of Parametric and Nonparametric Statistical Procedures . London, U.K.: Chapman & Hall, 2004. -
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
boostedcorrelational- 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.
It may state two separate claims as one.
- Also stated in: § VI. CONCLUSION · p. 13 · boosted
In the paper
“Notice that our approach produces better visual quality, and better objective quality, as gauged by the multi-scale structural similarity index (MS-SSIM) [29].”
What the sources say
- Not checked yetcites [29] [29] Z. Wang, E. P. Simoncelli, and A. C. Bovik, 'Multiscale structural similarity for image quality assessment,' in Proc. Asilomar Conf. Signals, Syst. Comput. , vol. 2. 2003, pp. 1398-1402. -
plaindirect- 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.
It may state two separate claims as one.
In the paper
“Our analysis indicates that for all noise variances simulated in the present data, our approach is statistically superior to the reference BM3D implementation in terms of perceived visual quality at the 95% confidence level, excepting when the noise variance is a tiny 0.0316 - where the two approaches become statistically indistinguishable.”
What the sources say
- Not checked yetcites [55] [55] L. V. Hedges and I. Olkin, Statistical Methods for Meta-Analysis . New York: Academic, 1985. <!-- image --> Anish Mittal received the B.Tech. degree in electrical engineering from the Indian Institute of Technology Roorkee, Roorkee, India, and the M.S. degree in computer engineering from The University of Texas at Austin, Austin, in 2009 and 2011, respectively. He joined the Laboratory for Image and Video Engineering, The University of Texas at Austin, in 2009, where he has been an Assistant Director since January 2012. His research interests include image and video processing, machine le
hedged- 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.
It may state two separate claims as one.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“JP2K and JPEG are also confused sometimes.”
hedgedcorrelational- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
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
“Having demonstrated that BRISQUE performs well in terms of correlation with human perception, we also now show that BRISQUE has low complexity.”
- Also stated in: § VI. CONCLUSION · p. 13 · boosted
“Further, we also showed that BRISQUE is computationally efficient and that its efficiency is superior to other distortion-generic approaches to NR IQA, thus making BRISQUE an attractive option for practical applications like image denoising.”
boosted- 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.
It may state two separate claims as one.
- Also stated in: § VI. CONCLUSION · p. 13 · boosted
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
“As Table X demonstrates, BRISQUE is quite efficient, outperforming the DIIVINE index and the BLIINDS-II index by a large amount.”
boosted- 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.
It may state two separate claims as one.
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
“As observed from the Table IV, the performance of BRISQUE remains relatively stable with respect to variation in the window size used to compute the local mean and variances. However, the performance starts to decrease when it becomes fairly large as the computations become non-local.”
plaindirect- 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.
It may state two separate claims as one.
In the paper
“Each of the distorted images has an associated difference mean opinion score (DMOS) which represents the subjective quality of the image.”
What the sources say
- Not matchedcites [27] [27] H. R. Sheikh, M. F. Sabir, and A. C. Bovik, 'A statistical evaluation of recent full reference image quality assessment algorithms,' IEEE Trans. Image Process. , vol. 15, no. 11, pp. 3440-3451, Nov. 2006. -
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
plaincorrelational- 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.
Probably not a substantial claim.
- Not matched
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“BRISQUE features may be used for distortion-identification as well.”
hedgedconditional- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“The product of our research is the Blind/Referenceless Image Spatial QUality Evaluator (BRISQUE) which utilizes an NSS model framework of locally normalized luminance coefficients and quantifies 'naturalness' using the parameters of the model.”
plaindirect- 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.
It may state two separate claims as one.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“The performance of any IQA model is best gauged by its correlation with human subjective judgements of quality, since the human is the ultimate receiver of the visual signal.”
plaincorrelational- 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.
Probably not a substantial claim.
What the paper did — it cites no source for this, so there is none to check. The quote shows it is in the paper.
In the paper
“Further, we repeat this random train-test procedure 1000 times and report the median of the performance across these 1000 iterations, in order to eliminate performance bias.”
plaindirect- 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.
Probably not a substantial claim. It may state two separate claims as one.
None of the sources cited for this claim have been checked yet.
In the paper
“The regularity of natural scene statistics (NSS) has been well established in the visual science literature, where regularity has been demonstrated in the spatial domain [15], and in the wavelet domain [16].”
What the sources say
- Not checked yetcites [15] [15] D. L. Ruderman, 'The statistics of natural images,' Netw. Comput. Neural Syst. , vol. 5, no. 4, pp. 517-548, 1994. -
- Not checked yetcites [16] [16] A. Srivastava, A. B. Lee, E. P. Simoncelli, and S. C. Zhu, 'On advances in statistical modeling of natural images,' J. Math. Imag. Vis. , vol. 18, no. 1, pp. 17-33, 2003. -
boosted- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Our claim is that characterizing locally normalized luminance coefficients in this way is sufficient not only to quantify naturalness, but also to quantify quality in the presence of distortion.”
plaindirect- 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: unclear — It is unclear why the paper says it holds.
It may state two separate claims as one.
None of the sources cited for this claim have been checked yet.
In the paper
“Although PSNR is a poor measure of perceptual quality, it is often used to benchmark for QA algorithms [41], [42].”
What the sources say
- Not checked yetcites [41] [41] Z. Wang and A. C. Bovik, 'Mean squared error: Love it or leave it? A new look at signal fidelity measures,' IEEE Signal Process. Mag. , vol. 26, no. 1, pp. 98-117, Jan. 2009. - source
- Not checked yetcites [42] [42] B. Girod, 'What's wrong with mean squared error?' in Visual Factors of Electronic Image Communications , A. B. Watson Ed. Cambridge, MA: MIT Press, 1993, pp. 207-220. -
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
What the paper did — it cites no source for this, so there is none to check. The quote shows it is in the paper.
In the paper
“we divide the LIVE database into two randomly chosen subsets - 80% training and 20% testing - such that no overlap between train and test content occurs.”
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.
It may state two separate claims as one.
None of the sources cited for this claim have been checked yet.
In the paper
“For example, it is well known that the power spectrum of natural images is a function of frequency and takes the form 1 / f γ , where γ is an exponent that varies over a small range across natural images.”
What the sources say
- Not checked yetcites [15] [15] D. L. Ruderman, 'The statistics of natural images,' Netw. Comput. Neural Syst. , vol. 5, no. 4, pp. 517-548, 1994. -
- Not checked yetcites [16] [16] A. Srivastava, A. B. Lee, E. P. Simoncelli, and S. C. Zhu, 'On advances in statistical modeling of natural images,' J. Math. Imag. Vis. , vol. 18, no. 1, pp. 17-33, 2003. -
boostedcorrelational- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. It may state two separate claims as one.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“To illustrate a new practical application of BRISQUE, we describe how a nonblind image denoising algorithm can be augmented with BRISQUE in order to perform blind image denoising.”
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
In the paper
“We used the LIVE IQA database [27] to test the performance of BRISQUE, which consists of 29 reference images with 779 distorted images spanning five different distortion categories - JPEG2000 (JP2K) and JPEG compression, additive white Gaussian noise (WN), Gaussian blur (Blur), and a Rayleigh fast-fading channel simulation (FF).”
What the sources say
- Not matchedcites [27] [27] H. R. Sheikh, M. F. Sabir, and A. C. Bovik, 'A statistical evaluation of recent full reference image quality assessment algorithms,' IEEE Trans. Image Process. , vol. 15, no. 11, pp. 3440-3451, Nov. 2006. -
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
- Not matched
A conclusion the paper draws from its own results — 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 slight dip in the performance can be attributed to imperfect distortion identification in the first stage of the two-stage framework.”
plaindirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
The paper states this without a source or data behind it here, so there is nothing to check it against. The quote shows it is in the paper.
In the paper
“Our contribution here is a new NR IQA model that is purely spatial; that relies on a spatial NSS model which does not require a mapping to a different co-ordinate domain (wavelet, DCT, etc.) and so is 'transform-free';”
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.
It may state two separate claims as one.
None of the sources cited for this claim have been checked yet.
In the paper
“Although image denoising is a well studied problem in image processing [50]-[54], blind image denoising remains relatively underexplored [49], [54].”
What the sources say
- Not checked yetcites [49] [49] X. Zhu and P. Milanfar, 'Automatic parameter selection for denoising algorithms using a no-reference measure of image content,' IEEE Trans. Image Process. , vol. 19, no. 12, pp. 3116-3132, Dec. 2010. -
- Not checked yetcites [54] [54] K. Dabov, A. Foi, V. Katkovnik, and K. Egiazarian, 'Image denoising by sparse 3-D transform-domain collaborative filtering,' IEEE Trans. Image Process. , vol. 16, no. 8, pp. 2080-2095, Aug. 2007. -
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: passes — The paper gives a reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it.
In the paper
“The denoising algorithm that we use is the one proposed in [54], which requires as input the noise variance in the image.”
What the sources say
- Not checked yetcites [54] [54] K. Dabov, A. Foi, V. Katkovnik, and K. Egiazarian, 'Image denoising by sparse 3-D transform-domain collaborative filtering,' IEEE Trans. Image Process. , vol. 16, no. 8, pp. 2080-2095, Aug. 2007. -
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.
Probably not a substantial claim. It may state two separate claims as one.
What the paper did — it cites no source for this, so there is none to check. The quote shows it is in the paper.
In the paper
“1000 images were then used for training and 2000 for testing thereby ensuring no content overlap between the two sets.”
plaindirect- 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.
It may state two separate claims as one.
In the paper
“The TID database consists of 25 reference images and 1700 distorted images over 17 distortion categories [4].”
What the sources say
- Not matchedcites [4] [4] N. Ponomarenko, V. Lukin, A. Zelensky, K. Egiazarian, M. Carli, and F. Battisti, 'TID2008-a database for evaluation of full-reference visual quality assessment metrics,' Adv. Modern Radioelectron. , vol. 10, no. 10, pp. 30-45, 2009. -
Couldn’t be matched to a paper in Crossref, so it can’t be checked.
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.
Probably not a substantial claim.
- Not matched
What the paper did — it cites no source for this, so there is none to check. The quote shows it is in the paper.
In the paper
“We use unoptimized MATLAB code for all of these algorithms in order to ensure a fair comparison.”
plaindirect- 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.
Probably not a substantial claim.
A conclusion the paper draws from its own results — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“Notice that BRISQUE augmentation is not limited to the BM3D algorithm; and any non-blind algorithm could be improved by using BRISQUE natural scene features to produce a blind image denoiser.”
hedgeddirect- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: fails — The paper’s conclusions may not depend on it.
- Based: fails — The paper may give no reason why it holds.
Probably not a substantial claim. The paper’s conclusions may not depend on it. The paper may give no reason why it holds. It may state two separate claims as one.
- Rests on an assumption: Existing blind image denoising algorithms typically address parameter estimation in an ad-hoc fashion without regard to natural scene statistics.
A conclusion the paper draws from its own results — 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 computational efficiency and excellent quality prediction performance makes BRISQUE an attractive option for practical applications.”
- Also stated in: § VI. CONCLUSION · p. 13 · boosted
“Further, we also showed that BRISQUE is computationally efficient and that its efficiency is superior to other distortion-generic approaches to NR IQA, thus making BRISQUE an attractive option for practical applications like image denoising.”
boosted- 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.
Probably not a substantial claim.
- Also stated in: § VI. CONCLUSION · p. 13 · boosted
In the paper
“The work closest in concept to this approach is the one proposed in [49] where image content measures were used to predict the noise variance in the image, which was then used for image denoising; however the approach is computationally intensive and the measure of content in the image may not be the ideal measure to predict noise variance.”
What the sources say
- Not checked yetcites [49] [49] X. Zhu and P. Milanfar, 'Automatic parameter selection for denoising algorithms using a no-reference measure of image content,' IEEE Trans. Image Process. , vol. 19, no. 12, pp. 3116-3132, Dec. 2010. -
hedgedconditional- 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.
It may state two separate claims as one.
The authors state this as an assumption.
The MSCN coefficients are zero mean and unit variance.
“assuming the MSCN coeffficients are zero mean and unit variance”
The paper relies on this without support.
Distortions affect image structure across scales.
“Images are naturally multiscale, and distortions affect image structure across scales.”
The paper relies on this without support.
MSCN coefficient distributions are symmetric.
“We choose the zero mean distribution, since (generally) MSCN coefficient distributions are symmetric.”
The paper relies on this without support.
Existing blind image denoising algorithms typically address parameter estimation in an ad-hoc fashion without regard to natural scene statistics.
“The proposed algorithms typically address parameter estimation in an ad-hoc fashion without regard to natural scene statistics.”
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
Each contradiction between two passages listed here was confirmed by an independent check.
Rated major: it would touch a result — a value a result rests on, reported and recomputed p-values on opposite sides of the significance level, two passages that conflict with no sign which is right, or a cited work retracted before this paper was published.
In the paper
§ C. Statistical Significance and Hypothesis Testing · p. 10
“| MSSSIM | 1 | 1 | 0 | 1 | 1 | 1 | 1 | 1 |”
§ C. Statistical Significance and Hypothesis Testing · p. 10
“| BRISQUE | 1 | 1 | 1 | 1 | 1 | 1 | 1 | 0 |”
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ G. Computational Complexity · p. 11
“parameter estimation needs to be performed only 5 times for an entire image”
§ A. Natural Scene Statistics in the Spatial Domain · p. 6
“Hence, we extract all features listed in Table I at two scales -the original image scale, and at a reduced resolution (low pass filtered and downsampled by a factor of 2).”
Rated minor: it would not touch a result — a value no result rests on, a p-value that stays on the same side of the significance level, or a conflict that is clerical, resolved elsewhere in the paper, or outside its own findings.
In the paper
§ C. Statistical Significance and Hypothesis Testing · p. 10
“| PSNR | 0 | - 1 | - 1 | - 1 | - 1 | 1 | - 1 | - 1 |”
§ C. Statistical Significance and Hypothesis Testing · p. 10
“| BLIINDS-II | 1 | - 1 | - 1 | 1 | 1 | 0 | - 1 | - 1 |”
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-citationLIBSVM: A library for support vector machines
Chang, C.-C., & Lin, C.-J. (2011) · ACM Transactions on Intelligent Systems and Technology · doi:10.1145/1961189.1961199 · Cited by 40,918
A Tutorial on Support Vector Machines for Pattern Recognition
Burges, C. J. C. (1998) · Data Mining and Knowledge Discovery · doi:10.1023/a:1009715923555 · Cited by 16,615
Wasserman, S. S., Hedges, L. V., & Olkin, I. (1988) · Journal of Educational Statistics · doi:10.2307/1164953 · Cited by 10,361
Libsvm : A library for support vector machines
Ferrari, V. (2008) · Medical Entomology and Zoology · Cited by 10,123
Image denoising by sparse 3-D transform-domain collaborative filtering
Dabov, K., Foi, A., Katkovnik, V., & Egiazarian, K. (2007) · IEEE Transactions on Image Processing · doi:10.1109/tip.2007.901238 · Cited by 9,283
Martin, D., Fowlkes, C., Tal, D., & Malik, J. (2002) · Proceedings Eighth IEEE International Conference on Computer Vision. ICCV 2001 · doi:10.1109/iccv.2001.937655 · Cited by 8,216
A Non-Local Algorithm for Image Denoising
Buades, A., Coll, B., & Morel, J.-M. (2005) · 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05) · doi:10.1109/cvpr.2005.38 · Cited by 7,616
Multiscale structural similarity for image quality assessment
Wang, Z., Simoncelli, E. P., & Bovik, A. C. (2004) · The Thrity-Seventh Asilomar Conference on Signals, Systems & Computers, 2003 · doi:10.1109/acssc.2003.1292216 · Cited by 6,253
Self-citationImage Denoising Via Sparse and Redundant Representations Over Learned Dictionaries
Elad, M., & Aharon, M. (2006) · IEEE Transactions on Image Processing · doi:10.1109/tip.2006.881969 · Cited by 5,321
Research works printed but not linked: 7
- wainwright, 2002, Statist Theories of the Brain, 203
- 10.1088/0954-898x/5/4/006 · DOI
- saad, 2012, BLIINDS Code
- moorthy, 2011, DIIVINE Code Implementation
- girod, 1993, Visual Factors of Electronic Image Communications, 207
- wang, 2003, MS-SSIM Code
- wang, 2004, SSIM Code
Cited by
Sorted by citations
Making a “Completely Blind” Image Quality Analyzer
Mittal, A., Soundararajan, R., & Bovik, A. C. (2013) · IEEE Signal Processing Letters · doi:10.1109/lsp.2012.2227726 · Cited by 6,766
Self-citationBenchmarking Single-Image Dehazing and Beyond
Li, B., Ren, W., Fu, D., Tao, D., Feng, D., Zeng, W., & Wang, Z. (2018) · IEEE Transactions on Image Processing · doi:10.1109/tip.2018.2867951 · Cited by 2,451
A Feature-Enriched Completely Blind Image Quality Evaluator
Zhang, L., Zhang, L., & Bovik, A. C. (2015) · IEEE Transactions on Image Processing · doi:10.1109/tip.2015.2426416 · Cited by 1,311
Self-citationConvolutional Neural Networks for No-Reference Image Quality Assessment
Kang, L., Ye, P., Li, Y., & Doermann, D. (2014) · 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/cvpr.2014.224 · Cited by 1,309
Deep Neural Networks for No-Reference and Full-Reference Image Quality Assessment
Bosse, S., Maniry, D., Muller, K.-R., Wiegand, T., & Samek, W. (2017) · IEEE Transactions on Image Processing · doi:10.1109/tip.2017.2760518 · Cited by 1,325
Metadata
- Authors
- Mittal, A., Moorthy, A. K., & Bovik, A. C.
- Affiliations
- The University of Texas at Austin (US)
- ORCID
- Bovik, A. C.
- Venue
- IEEE Transactions on Image Processing
- Published
- 2012-12
- Volume
- 21
- Issue
- 12
- Pages
- 4695-4708
- ISSN
- 1057-7149 (ISSN-L)
- 1941-0042
- OpenAlex
- W1982471090
- Semantic Scholar
- a2cad4e4fd946adf6cc87e483b2ba18579de1264
- Cited by
- 6,064
- References
- 56
- Citations by year
- 2026: 652 · 2025: 861 · 2024: 758 · 2023: 707 · 2022: 645 · 2021: 490 · 2020: 426 · 2019: 387 · 2018: 334 · 2017: 268 · 2016: 208 · 2015: 151 · 2014: 129 · 2013: 42 · 2012: 5
- FWCI
- 107.73
- Type
- journal article
- Language
- en
- Open access
- closed
- Locations
- IEEE Transactions on Image Processing (publishedVersion)
- PubMed (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
- no-reference image quality assessment, BRISQUE, natural scene statistics, spatial domain features, blind image denoising
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 2026-10-09, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-09 (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
