Multi-Kernel Filtering for Nonstationary Noise: An Extension of Bilateral Filtering Using Image Context

Dinggang Shen; Feihong Liu; Jun Feng; Pew-Thian Yap

REVIEW 5 major objections 5 minor 24 references

Multi-Kernel Filtering for Nonstationary Noise: An Extension of Bilateral Filtering Using Image Context

T0 review · 5 major / 5 minor · reviewed 2026-08-14 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read The paper claims that replacing the single fixed range kernel of bilateral filtering with cluster-tree-derived kernels lets it handle nonstationary noise automatically.

desk verdict A genuinely new way to make bilateral filtering context-aware, but the central equation contradicts its own prose and the experiments are too thin to support the claims. read the letter →

arxiv 1908.06307 v4 pith:MTDI3SCM submitted 2019-08-17 cs.CV

Feihong Liu , Jun Feng , Pew-Thian Yap , Dinggang Shen This is my paper · ORCID

classification cs.CV

keywords bilateralfilteringmulti-kernelfilternonstationarynoiseimagedenoisinghierarchicalclusteringspatiallyvaryingadaptivestructuralsimilarity

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Bilateral filtering is a widely used denoising method, but its range kernel width is a free parameter that presumes one global noise level, so variable noise defeats it. The paper proposes a multi-kernel filter (MKF) that learns the kernels from the image itself: a hierarchical clustering algorithm partitions the noisy image into a cluster tree, and each leaf cluster, together with two ancestor clusters, sets a local range-kernel width at every pixel. The authors report that one fixed MKF configuration maintains near-linear error curves across noise levels 10 to 1000 on BSD300 and that MKF also handles spatially varying noise on BrainWeb MRI, outperforming BF, TV, and CF in both MAE and SSIM. The practical importance is that the filter's most sensitive parameter, the noise scale, no longer needs manual initialization per image.

What carries the argument

The load-bearing object is the cluster tree: a coarse-to-fine hierarchy of coherent pixel groups built by iterative two-class EM on intensity (similarity clustering) followed by connected-component isolation (proximity clustering). It supplies the image context. Each leaf cluster $C_{t,k}$ contributes one range kernel for the pixels it contains, and its two ancestor clusters $C_{t-1,k^*}$ and $C_{t-2,k^{**}}$ contribute the contextual scaling factor $\tau_{t,k} = \sqrt{\sigma_{t-1,k^*} \sigma_{t-2,k^{**}}}$. The kernel in Eq. (11), $w(\vec x, \vec \xi \mid \sigma_{t,k}, \tau_{t,k}) = \exp\big[-(\vec x - \vec \xi)^2 / (2 h_{\vec x}^2) - (I(\vec x)-I(\vec \xi))^2 / (2 \tau_{t,k} \sigma_{t,k}^2)\big]$, sets the effective range width pixel by pixel. Thus the same local intensity gradient is smoothed heavily inside low-salience regions and preserved near salient boundaries; this is what makes the single fixed kernel of BF into multiple automatically learned, spatially varying kernels.

What would settle it

Take a clean image containing a known flat region and a sharp edge, add Gaussian noise of known standard deviation, run MKF's clustering, and record the leaf-cluster variances $\sigma_{t,k}$ for flat-region and edge-region clusters. If the estimated variance rises with edge contrast rather than matching the injected noise level, the kernel widths are tracking image structure, not noise, and the adaptivity claim fails.

Watch

Extended reading notes

Core claim

Bilateral filtering rests on one manually chosen range-kernel width $h_I$, which encodes a global noise level; when the noise level changes over the image or across images, that single width can be either too small to smooth noisy pixels or too large to preserve edges. The paper's core claim is that this single kernel should be replaced by many kernels whose widths are learned from the image itself. MKF builds a cluster tree from the noisy input, with each leaf cluster characterized by an estimated mean $\mu_{t,k}$ and standard deviation $\sigma_{t,k}$, and with the variances of two ancestor clusters entering through $\tau_{t,k} = \sqrt{\sigma_{t-1,k^*} \sigma_{t-2,k^{**}}}$. The effective range width at a pixel is then $\tau_{t,k} \sigma_{t,k}$. With this construction, the paper reports that on BSD300 with noise levels 10 to 1000, and on BrainWeb MRI with spatially varying Gaussian noise, MKF outperforms BF, TV, and CF on both MAE and SSIM.

Load-bearing premise

The load-bearing premise is that the intensity spread estimated from each small cluster of the noisy image measures local noise level rather than the scene's edges or texture; if the clustering encodes structure, then every adapted kernel width is off, and the filter's adaptivity no longer means what the paper claims.

Editorial extensions

If this is right

With one fixed configuration, MKF produces near-linear MAE and SSIM curves across noise levels 10 to 1000 on BSD300, whereas BF's curves bend sharply and depend on the hand-set kernel width.
On complex-valued BrainWeb MRI corrupted by spatially varying Gaussian noise, MKF yields lower MAE and higher SSIM than BF, TV, and CF, and its output keeps the phase-induced gradual transitions rather than producing mosaic blocks.
Because the clustering splits the image recursively with fixed rules, the method adapts the number of clusters to each image automatically, removing the manual membership count that flat clustering requires.
The two-ancestor context term makes the filter smooth across boundaries of low-salience leaf clusters while preserving boundaries of high-salience ones, which is the mechanism behind the reported texture preservation.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

Editorial inference: The cluster-tree conditioning is a general template; replacing the intensity-variance statistic with edge orientation or texture energy would give context-conditioned kernels for other bilateral-filter pipelines, such as tone mapping, depth upsampling, or edge-preserving smoothing.
Editorial inference: A direct attribution experiment is available: compare MKF against BF whose range width is set from the ground-truth noise-level map. If oracle BF matches MKF, the benefit reduces to per-pixel noise estimation; if it does not, the two-ancestor context term adds more than variance information.
Editorial inference: The Gestalt and visual-context narration is not required by the mathematics; the same equations can be read as a hierarchical mixture estimate of local noise scale, so the method can be adopted without committing to that framing.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Colleague,

I read the MKF paper. The core idea is genuine: replace BF's single range kernel with multiple kernels derived from a hierarchical cluster tree. Leaf cluster variance sets the range width per pixel, and ancestor clusters provide a contextual factor. That combination is new relative to the cited literature, and the clustering procedure—EM on intensity followed by proximity grouping—is described clearly enough to reproduce. The authors also frame a real problem: nonstationary noise in MRI and natural images.

The paper is not ready for publication, though, because of a load-bearing inconsistency. Eq. 11 puts tau_{t,k} in the denominator of the exponent, so the range Gaussian has variance tau * sigma^2. The prose in Section 3.2.2 says tau<1 "enlarges" sigma and that tau close to 0 allows smoothing across cluster boundaries. But tau<1 narrows the kernel, and tau->0 makes it a delta, preserving every intensity difference. So the mechanism as written does the opposite of what the text claims. The contextual factor is the only place where image context enters the filter, so this contradiction cuts to the heart of the method. If it's a typo, a corrected equation might make the method behave as intended; but a reader cannot verify which behavior was actually tested.

The experimental support is also thin. Only three baselines—BF, TV, CF—no error bars, no statistical tests, no code or data. Calling those "state-of-the-art" is generous. Figure 4 shows tree depth Dp must be tuned to noise level; the authors admit the performance depends on clustering parameters. That undercuts the "automatic" adaptivity claim in the abstract.

What's good: the idea is worth developing. The multi-scale cluster representation is a sensible way to make BF context-aware. The MRI experiments on complex data are a nice touch. But the current manuscript doesn't provide enough evidence to support the claims.

My recommendation: this deserves peer review, because the core idea is novel and the main flaw might be fixable. A competent referee could catch the equation inversion and demand a corrected version with proper baselines. I wouldn't accept it in this form, but I wouldn't bury it either.

For you: if you work on adaptive filtering, it's worth a skim. The cluster-tree construction might be reusable even if the noise model isn't.

Referee Report

5 major / 5 minor

Summary. The paper proposes Multi-Kernel Filter (MKF), an extension of bilateral filtering designed for nonstationary noise. It constructs a hierarchical cluster tree of image pixels via EM-based similarity clustering and proximity clustering, then uses leaf-cluster standard deviations and ratios of predecessor-cluster standard deviations to set spatially varying range-kernel widths. The authors claim that MKF adapts filtering kernels to image content automatically and report experiments on BSD300 and BrainWeb showing lower MAE and higher SSIM than BF, TV, and CF.

Significance. If the mechanism were correctly specified, the idea of using a cluster-tree context to set per-pixel range kernels would be a plausible contribution to adaptive image filtering, and the qualitative kernel illustrations in Figure 5 are suggestive. However, the central formula in Eq. (11) contradicts the accompanying prose, and the experimental evaluation is too thin to support the 'state-of-the-art' claim. The method's adaptivity is shown in Figure 4 to depend strongly on a manually chosen tree depth, and the paper does not provide code, error bars, or statistical tests. The idea is promising but the current manuscript does not establish its validity.

major comments (5)

[§3.2.2, Eq. (11)-(12)] The contextual factor as written does the opposite of what the text claims. In Eq. (11), the range-kernel denominator is 2 τ_{t,k} σ_{t,k}^2; since the paper states that σ_{t−2,k**} > σ_{t−1,k*}, Eq. (12) yields τ_{t,k} < 1, so the effective range variance τ_{t,k} σ_{t,k}^2 is smaller than σ_{t,k}^2. As τ_{t,k} → 0, the kernel approaches a delta in intensity, which blocks smoothing across intensity differences. The text in §3.2.2 instead states that τ close to 0 'significantly enlarges σ_{t,k}' and 'smooths across the boundaries' of C_{t,k}. This is an internal contradiction in the central mechanism, and the reader cannot determine whether Eq. (11) or the prose describes the intended behavior.
[§4.3, Fig. 4] Figure 4 shows that the optimal tree depth Dp depends strongly on the noise level: for noise level 10, t=7 gives the best MAE/SSIM, while for noise level 1000, t=7 gives the worst. Since Dp is a manually initialized parameter (§4.1), the method does not automatically adapt to noise nonstationarity; a user must know the noise level to choose Dp. This contradicts the central claim of automatic adaptivity and is acknowledged in the conclusion ('its performance depended on the parameters of the clustering'). The near-linear curves for t=2 in Fig. 6 are achieved by fixing Dp=2, not by adaptation.
[§4.2-4.4] The experimental evidence is too limited to support the claim that MKF 'outperforms state-of-the-art filters.' On BSD300, MKF is compared only with BF; on BrainWeb, it is compared with BF, TV, and CF, all from 1998, 1992, and 2017 respectively, which are not state-of-the-art denoisers for the reported tasks. No error bars, standard deviations, or statistical significance tests are reported. Also, no sensitivity analysis is given for baseline parameters beyond the statement that they were 'well chosen.' These omissions leave the quantitative superiority claim unsubstantiated.
[§3.1.1-§3.2] The adaptivity of MKF rests on the assumption that the leaf-cluster standard deviation σ_{t,k}, estimated by EM from the noisy image (Eqs. (2)-(3)), reflects local noise statistics rather than image structure. In natural images, σ_{t,k} is a mixture of both, and using it directly as the range-kernel width may over-smooth textured regions or under-smooth noisy flat regions. No experiment validates this assumption, for example by comparing the estimated σ_{t,k} with the ground-truth noise level in the synthetic BrainWeb data.
[§4.3, Eq. (12)] The definition of τ_{t,k} is not reproducible for t=2. Eq. (12) references σ_{t−2,k**}, which does not exist for t=2, but §4.3 states that τ_{t,k} = (σ_{t,k}/σ_{t−1,k*})^2 when t=2. These two definitions disagree, and the manuscript does not explain how the root-layer case is handled. This ambiguity affects the parameter setting used in the main experiments.

minor comments (5)

[§1 affiliation] The affiliation line contains a typo: 'Universtiy' should be 'University'.
[§3.2.1, Eq. (4)] The parentheses are unbalanced in 'O(I(x)))'; one closing parenthesis should be removed.
[§4.3, second paragraph] The cross-reference 'In Figure 4.3' should read 'In Figure 4'.
[Abstract] The abstract duplicates a sentence: 'We first design...' appears twice, and the 'At first, we design...' sentence repeats the same idea verbatim.
[§4.1, parameter list] The name 'Conduction coefficient' for the spatial kernel variance h_x is nonstandard; consider using 'spatial kernel width' instead.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity found: MKF estimates kernel widths from the noisy image itself via EM clustering, but no equation reduces the output to a fitted target; the main issue is an internal inconsistency in Eqs. 11–12, which is a correctness concern, not circularity.

full rationale

The derivation chain is self-contained in the sense required by the circularity test. The adaptive range-kernel width sigma_{t,k} is estimated from the noisy input image by EM clustering (Eqs. 2–3), and the contextual factor tau_{t,k} is formed as a ratio of cluster standard deviations from coarser levels (Eq. 12). This is self-referencing in the ordinary filtering sense—the filter uses statistics of the image being filtered—but it is not a fit to the target outputs: no parameter is regressed against the reported MAE or SSIM values, and the denoised image is not a relabeling of the fitted quantities. The claimed 'outperforms state-of-the-art filters' is an empirical comparison on BSD300 and BrainWeb, not a consequence derived from the construction. Self-citations (e.g., Liu et al. 2017 for clustering inspiration; Chen et al. 2019 for spatially varying MRI noise) are background or related-work citations and are not load-bearing for the filtering mechanism. There is no imported uniqueness theorem, no ansatz smuggled in via citation, and no renaming of a known result as a new organization. The conclusion openly admits that performance depends on clustering parameters and that the method is slow; those are limitations but not circularity. One substantive concern is an internal inconsistency in the text around Eqs. 11–12: the prose states that sigma_{t-2} > sigma_{t-1}, hence tau_{t,k} < 1, which 'enlarges sigma_{t,k}', and that tau close to 0 permits smoothing across leaf-cluster boundaries. In Eq. 11, however, tau_{t,k} multiplies sigma_{t,k}^2 in the denominator of the range-kernel exponent, so tau < 1 narrows the effective range width and tau -> 0 turns the range kernel into a delta, preventing cross-boundary smoothing. This is a specification error that would affect reproducibility and correctness of the claimed adaptive behavior, but it is not a circular reduction of the result to its inputs; it does not raise the circularity score.

Assumptions & free parameters 4 free parameters · 4 assumptions · 0 invented entities

The central claim depends on several hand-set clustering parameters, most notably tree depth and cluster size, which the experiments show must be tuned to noise level. The method also assumes that Gaussian mixture statistics of the noisy image capture local noise properties. No new physical or conceptual entities are introduced.

free parameters (4)

Maximal tree depth Dp = 2 and 7
Sets depth of cluster hierarchy; best value depends on noise level (Fig. 4), so it is a tuned parameter, not a fixed scale.
Maximal cluster size Mxc = 20
Caps pixels per cluster; chosen by hand in experiments and affects MAE and SSIM.
Spatial kernel sigma h_x = 3 (cat experiment), unspecified elsewhere
Controls spatial attenuation, inherited from BF and set manually.
Minimal cluster size Mnc = 9 (threshold)
Smaller clusters inherit parent statistics; threshold set to 9 pixels by hand.

assumptions (4)

domain assumption Additive Gaussian noise model (Eq. 1).
The paper assumes observed image = clean image plus Gaussian noise, with noise possibly nonstationary.
domain assumption Leaf-cluster variance reflects local noise statistics rather than image structure.
Central to Eq. 11, where sigma_{t,k} sets the range-kernel width.
domain assumption EM clustering with two components per split yields a meaningful image context hierarchy.
The cluster tree in Section 3.1 is built from repeated two-way EM splits.
ad hoc to paper Gestalt grouping rules motivate the clustering design.
The biological inspiration is not a formal requirement for the algorithm's correctness, but the paper uses it to justify the clustering structure.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Multi-Kernel Filtering for Nonstationary Noise: An Extension of Bilateral Filtering Using Image Context." pith.science (2026). https://pith.science/paper/MTDI3SCM

@misc{pith2026190806307,
  author       = {Pith},
  title        = {Pith review of: Multi-Kernel Filtering for Nonstationary Noise: An Extension of Bilateral Filtering Using Image Context},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/MTDI3SCM}},
  note         = {Machine review of arXiv:1908.06307}
}

read the original abstract

Bilateral filtering (BF) is one of the most classical denoising filters, however, the manually initialized filtering kernel hampers its adaptivity across images with various characteristics. To deal with image variation (i.e., non-stationary noise), in this paper, we propose multi-kernel filter (MKF) which adapts filtering kernels to specific image characteristics automatically. The design of MKF takes inspiration from adaptive mechanisms of human vision that make full use of information in a visual context. More specifically, for simulating the visual context and its adaptive function, we construct the image context based on which we simulate the contextual impact on filtering kernels. We first design a hierarchically clustering algorithm to generate a hierarchy of large to small coherent image patches, organized as a cluster tree, so that obtain multi-scale image representation. The leaf cluster and corresponding predecessor clusters are used to generate one of multiple range kernels that are capable of catering to image variation. At first, we design a hierarchically clustering framework to generate a hierarchy of large to small coherent image patches that organized as a cluster tree, so that obtain multi-scale image representation, i.e., the image context. Next, a leaf cluster is used to generate one of the multiple kernels, and two corresponding predecessor clusters are used to fine-tune the adopted kernel. Ultimately, the single spatially-invariant kernel in BF becomes multiple spatially-varying ones. We evaluate MKF on two public datasets, BSD300 and BrainWeb which are added integrally-varying noise and spatially-varying noise, respectively. Extensive experiments show that MKF outperforms state-of-the-art filters w.r.t. both mean absolute error and structural similarity.

Figures

Figures reproduced from arXiv: 1908.06307 by the authors.

**Figure 1.** Gaussian distributions learned from two rounds of similarity clustering. (a) An eagle image selected from BSD300. (b) After the first round, all pixels are aggregated into two groups which intensity distributions are modeled by #1 and #2 Gaussian distributions respectively (denoted by dash-curves). Solid blue curves in (c) and (d) show exact intensity distributions of the two groups of pixels respectively. Each soli… view at source ↗

**Figure 2.** Label maps of intermediate clustering stages. 3.1.1. Two-Stage Clustering for Multi-Scale Image Representation The two-stage clustering comprises (i) intensity-based similarity clustering, which simulates the Gestalt similarity rule, and (ii) and the connectedness-based proximity clustering, which simulates the Gestalt proximity rule. At the first stage, similarity clustering employs expectation maximization (EM) cl… view at source ↗

**Figure 3.** Partial branches of a cluster tree. Each bounding box denotes a cluster. We can see a cluster in 2 layer contains high-contrast contents, such as the whole eagle; while a cluster in 3 layer only contains one leg which is lower-contrast. We specify a cluster in an upper layer as image context of corresponding leaf clusters. is denoted by, ̂(⃗) = { (⃗) ℎ , ⃗ ℎ⃗ } . (8) Thus, the filtering kernel can be written in the … view at source ↗

Figures from the paper (6 more)

**Figure 4.** Figure 4: While a large tree depth performs well on the image with small-level noise, a small tree depth is optimal to restore image from severe noise, demonstrating a negative correlation between the tree depth and noise level. To observe the adaptivity under a series of integr…

**Figure 5.** Figure 5: Severe noise (1000) significantly corrupts image contents as well as the filtering kernel of BF. Three typical noise-free and noisy image patches, shown in (a) and (d) respectively, are used to demonstrate such degeneration. The filtering kernels of BF are shown in (b)…

**Figure 6.** Figure 6: The curve flatness demonstrates the robustness of filtering. Due to the parameter learning, MKF adapted to noise variabilities better than BF. demonstrated mosaic blocks; and MKF generated smoother results which were closer to the noise-free ground truth. For quantitat…

**Figure 7.** Figure 7: Noise free and noisy synthetic MRI magnitude, phase map, and the complex components. Real part Imaginary part Noise Free Noisy BF TV CF MKF −3000 3000 [PITH_FULL_IMAGE:figures/full_fig_p010_7.png]

**Figure 8.** Figure 8: Both real and imaginary images demonstrate promising denoising performance of MKF. tigating and simulating the adaptive visual mechanisms in designing a smoothing filter. However, some limitations are worth noting. Although MKF adaptively generated parameters of filter…

**Figure 9.** Figure 9: MKF achieves better MAE and SSIM indices than BF, TV, and CF. Gao, Z., Gao, Q., Tang, N., Shui, R., Shen, M., 2016. Organization principles in visual working memory: Evidence from sequential stimulus display. Cognition 146, 277–288. Gong, Y., Sbalzarini, I.F., 2017. Cu…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

24 extracted references · 24 canonical work pages

[1]

Spatially variant noise estimation in MRI: A homomorphic approach

Aja-Fernández, S., Pie, T., Vegas-Sánchez-Ferrero, G., et al., 2015. Spatially variant noise estimation in MRI: A homomorphic approach. Medical Image Analysis (MIA) 20, 184–197

work page 2015
[2]

Visual objects in context

Bar, M., 2004. Visual objects in context. Nature Reviews Neuroscience 5, 617–629

work page 2004
[3]

Fundamental relationship between bilateral ﬁltering, adaptive smoothing, and the nonlinear diffusion equation

Barash, D., 2002. Fundamental relationship between bilateral ﬁltering, adaptive smoothing, and the nonlinear diffusion equation. IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI) 24, 844–847

work page 2002
[4]

Fast ﬁlter transform for image processing

Burt, P.J., 1981. Fast ﬁlter transform for image processing. Computer Graphics and Image Processing 16, 20–51

work page 1981
[5]

Denoising of infant diffusion MRI data via graph framelet matching in x-q space

Chen, G., Dong, B., Zhang, Y ., Lin, W., Shen, D., Yap, P.T., 2019. Denoising of infant diffusion MRI data via graph framelet matching in x-q space. IEEE Transactions on Medical Imaging (TMI)

work page 2019
[6]

Real diffusion-weighted MRI enabling true signal averaging and increased diffusion contrast

Eichner, C., Cauley, S.F., Cohen-Adad, J., Möller, H.E., Turner, R., Setsompop, K., Wald, L.L., 2015. Real diffusion-weighted MRI enabling true signal averaging and increased diffusion contrast. NeuroImage 122, 373–384

work page 2015
[7]

Neurocognitive architecture of working memory.Neuron 88, 33–46

Eriksson, J., V ogel, E.K., Lansner, A., Bergström, F., Nyberg, L., 2015. Neurocognitive architecture of working memory.Neuron 88, 33–46

work page 2015
[8]

The perceptual root of object-based storage: An interactive model of perception and visual working memory

Gao, T., Gao, Z., Li, J., Sun, Z., Shen, M., 2011. The perceptual root of object-based storage: An interactive model of perception and visual working memory. Journal of Experimental Psychology: Human Perception and Performance 37, 1803–1823. Feihong Liu et al.: Preprint submitted to Elsevier Page 10 of 11 MKF for Nonstationary Noise: An extension of BF us...

work page 2011

Show all 24 references

[9]

Organization principles in visual working memory: Evidence from sequential stimulus display

Gao, Z., Gao, Q., Tang, N., Shui, R., Shen, M., 2016. Organization principles in visual working memory: Evidence from sequential stimulus display. Cognition 146, 277–288

2016
[10]

Curvature ﬁlters efﬁciently reduce certain variational energies

Gong, Y ., Sbalzarini, I.F., 2017. Curvature ﬁlters efﬁciently reduce certain variational energies. IEEE Transactions on Image Processing (TIP) 26, 1786–1798

2017
[11]

MRI simulation-based evaluation of image-processing and classiﬁcation methods

Kwan, R.S., Evans, A.C., Pike, G.B., 1999. MRI simulation-based evaluation of image-processing and classiﬁcation methods. IEEE Transactions on Medical Imaging (TMI) 18, 1085–1097

1999
[12]

Normalized euclidean super-pixels for medical image segmentation, in: International Conference on Intelligent Computing, Springer

Liu, F., Feng, J., Su, W., Lv, Z., Xiao, F., Qiu, S., 2017. Normalized euclidean super-pixels for medical image segmentation, in: International Conference on Intelligent Computing, Springer. pp. 586–597

2017
[13]

Revealing detail along the visual hierarchy: Neural clustering preserves acuity from V1 to V4

Lu, Y ., Yin, J., Chen, Z., Gong, H., Liu, Y ., Qian, L., Li, X., Liu, R., Andolina, I.M., Wang, W., 2018. Revealing detail along the visual hierarchy: Neural clustering preserves acuity from V1 to V4. Neuron 98, 417–428

2018
[14]

Martin, D., Fowlkes, C., Tal, D., Malik, J., 2001. A database of human segmented natural images and its application to evaluating segmentation algorithms and measuring ecological statistics, in: Proceedings of the IEEE International Conference on Computer Vision (ICCV), pp. 416–423

2001
[15]

Edge preserving smoothing

Nagao, M., Matsuyama, T., 1979. Edge preserving smoothing. Computer Graphics and Image Processing 9, 394–407

1979
[16]

Scale-space and edge detection using anisotropic diffusion

Perona, P., Malik, J., 1990. Scale-space and edge detection using anisotropic diffusion. IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI) 12, 629–639

1990
[17]

The Gestalt principle of similarity beneﬁts visual working memory

Peterson, D.J., Berryhill, M.E., 2013. The Gestalt principle of similarity beneﬁts visual working memory. Psychonomic Bulletin & Review 20, 1282–1289

2013
[18]

Noise ﬂoor removal via phase correction of complex diffusion-weighted images: Inﬂuence on DTI and Q-space metrics, in: MICCAI Workshop on Computational Diffusion MRI, pp

Pizzolato, M., Fick, R., Boutelier, T., Deriche, R., 2016. Noise ﬂoor removal via phase correction of complex diffusion-weighted images: Inﬂuence on DTI and Q-space metrics, in: MICCAI Workshop on Computational Diffusion MRI, pp. 21–34

2016
[19]

Nonlinear total variation based noise removal algorithms.Physica D: Nonlinear Phenomena 60, 259–268

Rudin, L.I., Osher, S., Fatemi, E., 1992. Nonlinear total variation based noise removal algorithms.Physica D: Nonlinear Phenomena 60, 259–268

1992
[20]

Normalized cuts and image segmentation

Shi, J., Malik, J., 2000. Normalized cuts and image segmentation. IEEE Transactions on Pattern Analysis and Machine Intelligence ( TPAMI) 22, 888–905

2000
[21]

Neuromodulation of attention

Thiele, A., Bellgrove, M.A., 2018. Neuromodulation of attention. Neuron 97, 769–785

2018
[22]

Bilateral ﬁltering for gray and color images, in: Proceedings of the IEEE International Conference on Computer Vision (ICCV), pp

Tomasi, C., Manduchi, R., 1998. Bilateral ﬁltering for gray and color images, in: Proceedings of the IEEE International Conference on Computer Vision (ICCV), pp. 839–846. Feihong Liu et al.: Preprint submitted to Elsevier Page 11 of 11 MKF for Nonstationary Noise: An extension...

1998
[23]

Denoising of diffusion MRI using random matrix theory

Veraart, J., Novikov, D.S., Christiaens, D., Ades-Aron, B., Sijbers, J., Fieremans, E., 2016. Denoising of diffusion MRI using random matrix theory. NeuroImage 142, 394–406

2016
[24]

Joint contour ﬁltering

Wei, X., Yang, Q., Gong, Y ., 2018. Joint contour ﬁltering. International Journal of Computer Vision (IJCV) 126, 1245–1265. Feihong Liu et al.: Preprint submitted to Elsevier Page 12 of 11

2018

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.

[1] [1]

Spatially variant noise estimation in MRI: A homomorphic approach

Aja-Fernández, S., Pie, T., Vegas-Sánchez-Ferrero, G., et al., 2015. Spatially variant noise estimation in MRI: A homomorphic approach. Medical Image Analysis (MIA) 20, 184–197

work page 2015

[2] [2]

Visual objects in context

Bar, M., 2004. Visual objects in context. Nature Reviews Neuroscience 5, 617–629

work page 2004

[3] [3]

Fundamental relationship between bilateral ﬁltering, adaptive smoothing, and the nonlinear diffusion equation

Barash, D., 2002. Fundamental relationship between bilateral ﬁltering, adaptive smoothing, and the nonlinear diffusion equation. IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI) 24, 844–847

work page 2002

[4] [4]

Fast ﬁlter transform for image processing

Burt, P.J., 1981. Fast ﬁlter transform for image processing. Computer Graphics and Image Processing 16, 20–51

work page 1981

[5] [5]

Denoising of infant diffusion MRI data via graph framelet matching in x-q space

Chen, G., Dong, B., Zhang, Y ., Lin, W., Shen, D., Yap, P.T., 2019. Denoising of infant diffusion MRI data via graph framelet matching in x-q space. IEEE Transactions on Medical Imaging (TMI)

work page 2019

[6] [6]

Real diffusion-weighted MRI enabling true signal averaging and increased diffusion contrast

Eichner, C., Cauley, S.F., Cohen-Adad, J., Möller, H.E., Turner, R., Setsompop, K., Wald, L.L., 2015. Real diffusion-weighted MRI enabling true signal averaging and increased diffusion contrast. NeuroImage 122, 373–384

work page 2015

[7] [7]

Neurocognitive architecture of working memory.Neuron 88, 33–46

Eriksson, J., V ogel, E.K., Lansner, A., Bergström, F., Nyberg, L., 2015. Neurocognitive architecture of working memory.Neuron 88, 33–46

work page 2015

[8] [8]

The perceptual root of object-based storage: An interactive model of perception and visual working memory

Gao, T., Gao, Z., Li, J., Sun, Z., Shen, M., 2011. The perceptual root of object-based storage: An interactive model of perception and visual working memory. Journal of Experimental Psychology: Human Perception and Performance 37, 1803–1823. Feihong Liu et al.: Preprint submitted to Elsevier Page 10 of 11 MKF for Nonstationary Noise: An extension of BF us...

work page 2011