Adaptive Structure-constrained Robust Latent Low-Rank Coding for Image Recovery

Lei Wang; Meng Wang; Sheng Li; Yang Wang; Zhao Zhang; Zhengjun Zha; Zheng Zhang

REVIEW 3 major objections 5 minor 43 references

Adaptive Structure-constrained Robust Latent Low-Rank Coding for Image Recovery

T0 review · 3 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 This paper claims that one objective can jointly learn low-rank reconstruction codes, a group-sparse projection for salient features, and an adaptive weighting matrix that forces the codes into block-diagonal form, and that the resulting…

desk verdict Plausible incremental model, but the reported superiority is unverifiable because parameters are tuned on the test sets and no code is released. read the letter →

arxiv 1908.07860 v2 pith:V247AEL2 submitted 2019-08-21 cs.CV

Zhao Zhang , Lei Wang , Sheng Li , Yang Wang , Zheng Zhang , Zhengjun Zha , Meng Wang This is my paper · ORCID

classification cs.CV

keywords low-rankrepresentationlatentsubspacerecoveryblock-diagonalstructureadaptiveweightinggroupsparsefeatureextractionimageL21-norminexactALM

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

The paper proposes AS-LRC, a low-rank coding model that decomposes data as a low-rank reconstruction, a group-sparse salient feature part, and a sparse error, while also learning an adaptive weighting matrix that constrains the reconstruction codes to be block-diagonal. The motivation is that previous latent low-rank methods either require a hand-chosen kernel width or neighborhood size to preserve local structure, or cannot handle data from multiple disjoint subspaces. AS-LRC replaces the hand-chosen weight with an auto-weighting matrix computed from the same salient features it is extracting, so the structure constraint adapts to the data. The paper reports higher accuracies than LatLRR, SA-LatLRR, FLLRR, rLRR and I-LSPFC across face, object, and handwriting databases, with the reported gap growing as the level of image corruption increases.

What carries the argument

The central object is the auto-weighting matrix $R$, computed from the locality-adaptive salient features $LX$, which enters the objective through the structure-constraint term $(ee^T - R) \odot Z$. A sum-to-one constraint on the columns of $R$, together with an $L_{2,1}$-norm that makes many rows of $R$ zero, turns $R$ into a soft structural prior that pushes $Z$ toward a block-diagonal form matching the data's subspaces. The singular value thresholding and $L_{2,1}$-shrinkage steps of the inexact ALM update scheme carry out the optimization, but the load-bearing idea is that the weighting is learned from the same features that define the recovery, rather than fixed in advance.

What would settle it

Run AS-LRC on synthetic data drawn from a known union of $k$ independent subspaces with known labels; if the recovered $Z$ is not block-diagonal as measured by, say, normalized mutual information between the affinity graph of $Z$ and the ground-truth labels, or if the converged point has a nonzero KKT residual for Eq. (9), then the central claim that the adaptive weight enforces block-diagonal structure fails.

Watch

Extended reading notes

Core claim

AS-LRC performs the latent decomposition $X \approx XZ + LX + E$, where $Z$ is the coefficient matrix whose block-diagonal structure is enforced by an adaptive weight matrix $R$, $L$ is a projection that yields group-sparse salient features, and $E$ is the sparse error. The model is a single minimization combining a nuclear-norm term on the low-rank codes, an $L_{2,1}$-norm on the projection and on $R$, and a Frobenius term $\|LX - LXR\|_F^2$ that ties the auto-weighting to the extracted features. The central claim is that by computing $R$ jointly with the representation, the codes become block-diagonal without the user choosing a kernel width or neighborhood size, and the $L_{2,1}$-regularized projection gives robustness to noise and outliers that nuclear-norm feature learning does not.

Load-bearing premise

The load-bearing premise is that minimizing a six-block, non-jointly-convex objective with an inexact ALM scheme actually reaches a useful solution; the paper notes that convergence guarantees for inexact ALM are only established for two blocks, so the reported optimum is not demonstrated to be a stationary point.

Editorial extensions

If this is right

If AS-LRC matches the reported results, hand-tuned graph weights in structure-constrained low-rank coding can be replaced by weights learned inside the objective, removing a meta-parameter and improving adaptability to different datasets.
The group-sparse projection $L^*$ provides an out-of-sample path: new data are classified by projecting onto $L^*$ and then applying the learned linear classifier, which the paper demonstrates across training-set sizes.
The reported robustness at high corruption rates implies the method can act as a preprocessing step for recognition on heavily corrupted face, object, and handwriting images.
The reported convergence in 50 to 150 iterations, faster than LatLRR and SC-LRR in the paper's experiments, suggests the approach is practically efficient despite the six-block alternating scheme.

Reading between the lines

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

Because $R$ is learned from the same projection that produces the salient features, the learned weights inherit whatever biases the projection carries; on datasets where the salient features are unreliable, the auto-weighting could reinforce errors rather than correct them.
The block-diagonal claim could be tested directly on synthetic data from a known union of independent subspaces with ground-truth labels; reporting normalized mutual information between the affinity graph of $Z$ and the true labels would separate the structure constraint's contribution from the rest of the recognition pipeline.
The six-block inexact ALM has no proven convergence guarantee, so a reformulation with fewer blocks, or with a jointly convex surrogate, would make the optimization theory apply and is a natural next step.
The $L_{2,1}$ choice trades nuclear-norm feature learning for group sparsity; on data where salient features are not naturally group-sparse, the representation $L^*X$ may lose information that a low-rank projection would preserve.

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

First, the headline: the paper is an incremental but coherent combination of existing low-rank coding ideas, and the specific joint adaptive weighting R based on salient features does appear to be new. But the central empirical claim—consistent superiority over SA-LatLRR, I-LSPFC, etc.—is currently unverifiable, because the authors tune α and β on the test sets and report the best results. That is textbook test-set selection, and with no code or parameter files, the reported margins (e.g., 75.57 vs 71.26 at 10 train on YaleB-UMIST) could be an artifact of extra tuning degrees of freedom rather than a property of the model. The stress-test note has this right.

What the paper does well: the model is cleanly motivated. Combining LatLRR's latent decomposition with SC-LRR's block-diagonal regularization, and replacing the predefined affinity matrix with an auto-weighting matrix learned from the salient features L*X, is a sensible idea. The use of L2,1-norm on the projection instead of nuclear norm for feature extraction is also reasonable. The optimization, while not jointly convex, is a standard six-block inexact ALM with the usual updates, and the paper honestly admits in Section III-B that convergence theory is only available for two blocks. The parameter sensitivity analysis in Fig. 9, done on a separate subsection, gives some evidence of stability, though it is still on the test set.

Where it is soft: the tuning protocol is the biggest problem. Section IV states the parameters are selected from a wide grid and 'the best results over tuned parameters are reported for evaluation.' If that means selecting on test accuracy, and there is no validation split described, the comparisons in Tables 2–4 and Figs. 4–7 are not informative about real generalization. Second, the convergence plots in Fig. 8 only show residuals of the objective being minimized; they do not establish that the fixed point is a stationary point of Eq. (9). That is a minor issue in practice because inexact ALM often works, but it should be stated as empirical, not theoretical. Third, the OCR text is garbled in places; some equations (e.g., Eq. (13)) are hard to verify, so a careful referee would need the clean source.

Who this is for: researchers working on low-rank subspace recovery and robust image representation will find the formulation interesting as a combination of existing building blocks. It would be a useful paper if the authors release code, specify a validation protocol that does not touch test labels, and redo the comparisons under those conditions.

Recommendation: send to peer review with the expectation of major revision. The idea has merit and the literature gap is real, but the empirical core needs to be re-run and auditable. I would not desk-reject it, but I would not accept it as is.

Referee Report

3 major / 5 minor

Summary. The manuscript proposes AS-LRC, a robust latent low-rank coding model that decomposes data X into a low-rank reconstruction XZ with an adaptive block-diagonal structure constraint, a group-sparse salient feature part LX, and a sparse error E. The projection L and the adaptive weight matrix R are learned jointly by minimizing the objective in Eq. (9), which is solved by a multi-block inexact Augmented Lagrange Multiplier procedure (Algorithm 1). The paper evaluates AS-LRC on handwriting, face, and object databases for image recovery, de-noising, and recognition, reporting advantages over PCA, IRPCA, LatLRR, SA-LatLRR, FLLRR, rLRR, and I-LSPFC. The manuscript also includes parameter sensitivity plots and convergence curves.

Significance. If the empirical claims are substantiated, AS-LRC would be a reasonably useful incremental contribution: it combines latent low-rank representation with an auto-weighted block-diagonal regularizer and an L2,1-norm projection in a single objective, and the experimental coverage across many datasets is broad. The paper also provides a detailed algorithmic description and pseudocode. However, the central value of the paper is the claimed performance improvement, and that claim currently rests on a tuning protocol that selects parameters on the test set. The convergence argument for the proposed optimization is also not established for a multi-block nonconvex objective. These are load-bearing issues that need to be repaired before the results can be accepted. The idea itself is plausible, so the appropriate path is a major revision rather than rejection.

major comments (3)

[Section IV (Experimental Results), Tables 2–4 and Fig. 9] The reported superiority of AS-LRC may be a test-set tuning artifact. The text states that the two model parameters alpha and beta are chosen from {10^-8, 10^-6, ..., 10^8} and that "the best results over tuned parameters are reported for evaluation." Since the same test accuracy is later used as the evaluation metric (Fig. 9 explicitly plots test-set accuracy as a function of parameters), AS-LRC is given two extra tuning degrees of freedom relative to the one-parameter baselines. The margins in Tables 2–4, e.g., 75.57 vs 71.26 for I-LSPFC at 10 train in Table 2, could therefore reflect selection on the test labels rather than a property of the model. The authors should either fix parameters on a validation split, report the full accuracy surface over the grid for at least one dataset, or provide a reproducible tuning protocol with code. Without this, the central empirical claim is not verifiable.
[Section III-B, Eq. (9) and Algorithm 1] The statement that the objective is "generally convex" is not correct for the formulated problem. Equation (9) contains products of unknown variables: L and R appear multiplied in ||LX - LXR||_F^2, and R is multiplied by Z in the structure term (ee^T - R) ⊙ Z, so the problem is not jointly convex. The authors correctly acknowledge that the convergence of inexact ALM is only established for two blocks, while Algorithm 1 updates more than two blocks. Consequently, there is no guarantee that the reported output is a stationary point of Eq. (9). The convergence plots in Fig. 8 show residuals of the same objective that is being minimized, which does not verify convergence to an optimum. The paper should either provide a formal convergence result under explicit assumptions, or explicitly characterize the algorithm as heuristic and add a more meaningful convergence check (e.g., relative objective change or stationarity measures).
[Reproducibility and baseline tuning] No code, parameter files, or exact per-entry parameter values are provided, and the description of baseline tuning as "carefully chosen" is too vague to audit. Since the claimed gains over the strongest baselines are often only a few percent (e.g., Table 3 at 10 train: 58.39 vs 56.95 for I-LSPFC), small tuning differences can change the conclusions. Please release the code and the chosen parameter values for every reported table entry, or at minimum provide the selected (alpha, beta, lambda) for each dataset and training size.

minor comments (5)

[Section IV-A] The text says all images are downsampled to 32x32 pixels, but the face recovery experiments use data matrices of size 128x192; please clarify the image sizes used in the visualization experiments.
[Table 2] In the FLLRR row, the Mean and Best entries for the 12-train setting are both 68.92, which appears to be a typographical error and should be corrected.
[Fig. 8] Please define the plotted "convergence error" explicitly (e.g., the primal residual used in Algorithm 1's convergence check) and state whether the curves are averaged over the ten runs mentioned in the text.
[Algorithm 1] The algorithm description says the complexity is the same as LRR and IRPCA, but the updates in Eqs. (12)–(14) include matrix inversions whose cost depends on the problem dimensions; please state the per-iteration complexity more precisely.
[Eq. (9)] The notation for the Hadamard product is inconsistent, and the vector e in ee^T is not defined in Eq. (9); please introduce e as the all-ones vector of appropriate size and use a single symbol for the elementwise product.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation: AS-LRC is an optimization objective with empirical evaluation; tuning and convergence caveats are correctness risks, not circularity.

full rationale

The paper's derivation chain is an optimization objective (Eqs. 8-10) solved by inexact ALM, followed by empirical comparisons on standard datasets. None of the outputs (Z, L, E, R) is defined in terms of the reported accuracies; accuracies are measured on test or corrupted samples only after the model is solved. The block-diagonal regularizer uses (ee^T - R)⊙Z with R learned inside the same objective, so no fitted parameter is later renamed as a prediction. Self-citations such as I-LSPFC [9] and SA-LatLRR [11] appear as baselines or related work, not as load-bearing premises, and no uniqueness or impossibility result is imported from the authors' prior work. The paper explicitly admits in Section III-B that convergence of inexact ALM is established only for two blocks and that proving convergence for more than two blocks is 'tricky'; this is an honest limitation of the optimization claim, not a circular step. Section IV reports that α and β are chosen from a grid and 'the best results over tuned parameters are reported for evaluation'; if tuning used test labels, this is a selection or reproducibility risk, but it is not circular because the model equations do not encode the reported accuracy. No equation reduces to itself by construction, and no central claim is justified solely by a self-citation chain, so the appropriate finding is no significant circularity.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The paper introduces no new physical or conceptual entities beyond an optimization model. Its load-bearing assumptions are the standard low-rank-plus-sparse data model, the effectiveness of a learned similarity matrix as a block-diagonal prior, and the practical convergence of a six-block alternating scheme that the authors admit lacks theoretical guarantees.

free parameters (3)

alpha = per dataset, grid-searched over {10^-8,...,10^8}; e.g., alpha=0.01 on AR in Fig. 9c
Controls the L2,1-norm group sparsity of the adaptive weighting matrix R; selected by grid search per dataset and the best test result is reported.
beta = per dataset, grid-searched over {10^-8,...,10^8}; e.g., beta=0.1 on AR in Fig. 9a
Controls the strength of the block-diagonal structure constraint (ee^T - R)⊙Z; selected by grid search per dataset and the best test result is reported.
lambda = per dataset; e.g., lambda=0.015 on AR in Fig. 9
Balances the sparsity of the error term E; chosen per dataset with the other parameters by grid search.

assumptions (4)

domain assumption Observed data X can be decomposed as XZ + LX + E with XZ low-rank and E sparse.
This LatLRR-style decomposition (Eq. (3) and Eq. (8)) assumes the data lie near a union of low-rank subspaces with corruptions; required for the recovery claim.
domain assumption The auto-weighting matrix R learned from salient features LX correctly encodes block or subspace membership so that (ee^T - R)⊙Z enforces block-diagonal codes.
The entire structure-constraint mechanism (Eqs. (8)-(9)) relies on R reflecting locality or class structure of the latent features; if it does not, the regularization misleads the low-rank coding.
ad hoc to paper The multi-block inexact ALM updates (Eqs. (12)-(21)) converge to a useful solution.
The paper states in Section III-B that convergence is only theoretically established for two blocks and is 'tricky' for more; Algorithm 1 uses six blocks. Empirical convergence plots (Fig. 8) are the only support.
standard math L2,1-norm induced group sparsity on L and R improves robustness to noise and outliers.
Standard property of group-sparse norms, used in prior work (I-LSPFC); not specific to AS-LRC.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Adaptive Structure-constrained Robust Latent Low-Rank Coding for Image Recovery." pith.science (2026). https://pith.science/paper/V247AEL2

@misc{pith2026190807860,
  author       = {Pith},
  title        = {Pith review of: Adaptive Structure-constrained Robust Latent Low-Rank Coding for Image Recovery},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/V247AEL2}},
  note         = {Machine review of arXiv:1908.07860}
}

read the original abstract

In this paper, we propose a robust representation learning model called Adaptive Structure-constrained Low-Rank Coding (AS-LRC) for the latent representation of data. To recover the underlying subspaces more accurately, AS-LRC seamlessly integrates an adaptive weighting based block-diagonal structure-constrained low-rank representation and the group sparse salient feature extraction into a unified framework. Specifically, AS-LRC performs the latent decomposition of given data into a low-rank reconstruction by a block-diagonal codes matrix, a group sparse locality-adaptive salient feature part and a sparse error part. To enforce the block-diagonal structures adaptive to different real datasets for the low-rank recovery, AS-LRC clearly computes an auto-weighting matrix based on the locality-adaptive features and multiplies by the low-rank coefficients for direct minimization at the same time. This encourages the codes to be block-diagonal and can avoid the tricky issue of choosing optimal neighborhood size or kernel width for the weight assignment, suffered in most local geometrical structures-preserving low-rank coding methods. In addition, our AS-LRC selects the L2,1-norm on the projection for extracting group sparse features rather than learning low-rank features by Nuclear-norm regularization, which can make learnt features robust to noise and outliers in samples, and can also make the feature coding process efficient. Extensive visualizations and numerical results demonstrate the effectiveness of our AS-LRC for image representation and recovery.

Figures

Figures reproduced from arXiv: 1908.07860 by the authors.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

43 extracted references · 41 canonical work pages

[1]

Robust principal component analysis?

E. J. Candes, X. Li, Y. Ma, and J. Wright, “Robust principal component analysis?” Journal of the ACM, vol. 58, no. 3, pp. 1-11, 2011

work page 2011
[2]

Robust principal component analysis: Exact recovery of corrupted low -rank matrices via convex optimization,

J. Wright, A. Ganesh, S. Rao, Y. Peng, and Y. Ma, “Robust principal component analysis: Exact recovery of corrupted low -rank matrices via convex optimization,” in Advances in neural information processing systems, pp. 2080-2088, 2009

work page 2009
[3]

Robust subspace segmentation by low -rank representation,

G. Liu, Z. Lin, and Y. Yu, “Robust subspace segmentation by low -rank representation,” in : Proceedings of the 27th international conference on machine learning (ICML-10), pp. 663-670, 2010

work page 2010
[4]

Structure -constrained low -rank representation,

K. Tang, R. Liu, Z. Su, and J. Zhang, “Structure -constrained low -rank representation,” IEEE Transactions on Neural Networks and Learning Systems, vol. 25, no. 12, pp. 2167-2179, 2014

work page 2014
[5]

Subspace clustering by bl ock diagonal representation,

C. Lu, J. Feng, Z. Lin, T. Mei, and S. Yan, “Subspace clustering by bl ock diagonal representation,” IEEE Trans. Pattern Anal. Mach. Intell., vol. 41, no. 2, pp. 487–501, 2019

work page 2019
[6]

Inductive robust principal component analysis,

B.-K. Bao, G. Liu, C. Xu, S. Yan, “Inductive robust principal component analysis,” IEEE Transactions on Image Processing , vol. 21, no. 8, pp. 3794–3800, 2012

work page 2012
[7]

Latent low -rank representation for subspace segmentation and feature extraction,

G. C. Liu and S. C. Yan, “Latent low -rank representation for subspace segmentation and feature extraction,” in: Proceedings of the International Conference on Computer Vision, pp. 1615–1622, 2011

work page 2011
[8]

Subspace clustering based on latent low rank representation with Frobenius norm minimization,

S. Yu and Y. Q. Wu, “Subspace clustering based on latent low rank representation with Frobenius norm minimization,” Neurocomputing , vol. 275, pp. 2479–2489, 2018

work page 2018

Show all 43 references

[9]

Joint low -rank and sparse principal feature coding for enhanced robust representation and visu al classiﬁcation,

Z. Zhang, F. Li, M. Zhao, L. Zhang, S. Yan, “Joint low -rank and sparse principal feature coding for enhanced robust representation and visu al classiﬁcation,” IEEE Transactions on Image Processing, vol. 25, no. 6, pp. 2429–2443, 2016

2016
[10]

Similarity preserving low -rank representation for enhanced data representation and effective subspace learning,

Z. Zhang, S. C. Yan, and M. B. Zhao, “Similarity preserving low -rank representation for enhanced data representation and effective subspace learning,” Neural Networks, vol. 53, pp. 81–94, 2014

2014
[11]

Similarity adaptive latent low -rank representation for robust data representation,

L. Wang, Z. Zhang, S. Li, G. Liu, C. Hou, and J. Qin, “Similarity adaptive latent low -rank representation for robust data representation,” in: Proceedings of the Paciﬁc Rim International Conference on Artiﬁcial Intelligence, Springer, pp. 71–84, 2018

2018
[12]

Laplacian regularized low-rank representation and its applications,

M. Yin, J. Gao, Z. Lin, “Laplacian regularized low-rank representation and its applications,” IEEE Trans. Pattern Anal. Mach. Intell. , vol. 38, no. 3, pp. 504–517, 2016

2016
[13]

The augmented lagrange multiple method for exact recovery of corrupted low -rank matrices,

Z. Lin, M. Chen, and Y. Ma, “The augmented lagrange multiple method for exact recovery of corrupted low -rank matrices,” arXiv preprint arXiv : 1009.5055, 2010

2010 arXiv
[14]

A singular value thresholding algorithm for matrix completion,

J. F. Cai, E. J. Cand es, Z. Shen, “A singular value thresholding algorithm for matrix completion,” SIAM Journal on Optimization , vol. 20, no. 4, pp. 1956–1982, 2010

1956
[15]

Optshrink: An algorithm for improved low -rank signal matrix denoising by optimal, data -driven singular value shrinkage,

R. R. Nadakuditi, “Optshrink: An algorithm for improved low -rank signal matrix denoising by optimal, data -driven singular value shrinkage,” IEEE Transactions on Information Theory , vol. 60, no. 5, pp. 3002-3018, 2014

2014
[16]

Unipen project of on-line data exchange and recognizer benchmarks,

I. Guyon, L. Schomaker, R. Plamondon, M. Liberman, and S. Janet, “Unipen project of on-line data exchange and recognizer benchmarks,” in Proceedings of IAPR International Conference on Pattern Recognition , IEEE, pp. 29-33, 1994

1994
[17]

Optimal linear representations of images for object recognition,

A. S. X. Liu and K. Gal livan, “Optimal linear representations of images for object recognition,” in : Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition , 2003

2003
[18]

Combining models and exemplars for face recognition: An illuminating example,

T. Sim and K. Takeo, "Combining models and exemplars for face recognition: An illuminating example, " in: Proceedings of the CVPR Workshop on Models versus Exemplars in Computer Vision , 2001

2001
[19]

Automatic classiﬁcation of single facial images,

J. B. M. J. Lyons and S. Akamatsu, “Automatic classiﬁcation of single facial images,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 21, no. 12, pp. 1357–1362, 1999

1999
[20]

Making a science of model search

J. Bergstra, Y. Dan and David D. Cox, “Making a science of model search.” arXiv preprint arXiv:1209.5111, 2012

2012 arXiv
[21]

The CMU pose illumination and expression database,

T. Sim, S. Baker, M. Bsat, “The CMU pose illumination and expression database,” IEEE Trans. on Pattern Analysis and Machine Intelligence, vol. 25, no. 12, pp. 1615–1618, 2003

2003
[22]

From few to many: Illumination cone models for face recognition under variable lighting and pose,

A. S. Georghiades, P. N. Belhumeur, and D. Kriegman, “From few to many: Illumination cone models for face recognition under variable lighting and pose,” IEEE Trans. Pattern Anal. Mach. Intell., vol. 23, no. 6, pp. 643-660, Jun. 2001

2001
[23]

Characterising virtual eigensignatures for general purpose face recognition,

D. B. Grahan and N.M. Allinson, “Characterising virtual eigensignatures for general purpose face recognition,” in Face Recognition: From Theory to Applications, vol. 163, pp. 446-456, 1998

1998
[24]

Learning generative visual models from few training examples: An incremental Bayesian approach tested on 101 object categories,

F. F. Li, R. Fergus, and P. Perona. “ Learning generative visual models from few training examples: An incremental Bayesian approach tested on 101 object categories, ” in: Proceedings of the IEEE Computer Society Conference on Computer Vision & Pattern Recogn ition Workshop, 2004

2004
[25]

A singular value thresholding algorithm for matrix completion,

J. F. Cai, E. J. Candes Z. Shen, “A singular value thresholding algorithm for matrix completion,” SIAM J. Optim, . Not. 20, no. 4, pp. 1956 -1982, 2010

1956
[26]

Robust subspace segmentation by self-representation constrained low -rank representation

W. Lai, X. F. Wang, A. Wu, R. Zhou and C. M. Zhu, “Robust subspace segmentation by self-representation constrained low -rank representation”, Neural Processing Letters, no. 11, pp. 1-21, 2018

2018
[27]

Optimization algorithms for the missing data problem in a low- rank matrix, Part II - Why is the Gauss -Newton Hessian matrix preferred in Newton-like methods?

P. Chen. “Optimization algorithms for the missing data problem in a low- rank matrix, Part II - Why is the Gauss -Newton Hessian matrix preferred in Newton-like methods?” International Journal of Computer Vision , vol. 80, no. 1, pp. 125-142, 2019

2019
[28]

Multiview spectral clustering via structured low -rank matrix factorization,

Y. Wang, L. Wu, X. Lin and J. Gao, “Multiview spectral clustering via structured low -rank matrix factorization,” IEEE Transactions on Neural Networks & Learning Systems , vol. PP, no. 99, pp. 1-11, 2018

2018
[29]

Joint Label Prediction based Semi -Supervised Adaptive Concept Facto rization for Robust Data Representation,

Z. Zhang, Y. Zhang, G. Liu, J. Tang, S.Yan and M. Wang, "Joint Label Prediction based Semi -Supervised Adaptive Concept Facto rization for Robust Data Representation," IEEE Transactions on Knowledge and Data Engineering, 2019

2019
[30]

Semi -Supervised Local Multi -Manifold Isomap by Linear Embedding for Feature Extraction,

Y. Zhang, Z. Zhang, J. Qin, L. Zhang, B. Li and F. Li, "Semi -Supervised Local Multi -Manifold Isomap by Linear Embedding for Feature Extraction," Pattern Recognition, vol.76, pp.662-678, 2018

2018
[31]

Matrix Completion Based on Non -Convex Low - Rank Approximation,

F. Nie, Z. Hu, X. Li, “Matrix Completion Based on Non -Convex Low - Rank Approximation,” IEEE Trans. on Image Processing , vol.28, no.5, pp.2378-2388, 2019

2019
[32]

Object Classification With J oint Projection and Low -Rank Dictionary Learning,

H. Foroughi, N. Ray, and H. Zhang, “Object Classification With J oint Projection and Low -Rank Dictionary Learning,” IEEE Trans. on Image Processing, vol.27, no.2, pp.806-821, 2018

2018
[33]

Adaptive Low -Rank Multi-Label Active Learning for Image Classification,

J. Wu, A. Guo, V. S. Sheng, P. Zhao, Z. Cui, H. Li, “Adaptive Low -Rank Multi-Label Active Learning for Image Classification,” in: Proceedings of the ACM International Conference on Multimedia , pp.1336-1344, 2017

2017
[34]

Deep Low -rank Sparse Collective Factorization for Cross -Domain Recommendation,

S. Jiang, Z. Ding, Y.Fu, “Deep Low -rank Sparse Collective Factorization for Cross -Domain Recommendation,” In: Proceedings of the 25th ACM International Conference on Multimedia, pp.163-171, 2017

2017
[35]

Analyzing appearance and contour based methods for object categorization ,

B. Leibe, B. Schiele, “Analyzing appearance and contour based methods for object categorization ,” in Proc. IEEE Conf. Comput. Vis. Pattern Recognit., Madison, WI, USA, pp.409-415, 2003

2003
[36]

Adaptive Neighborhood Propagation by Joint L2,1 -norm Regularized Sparse Coding for Representation and Classification,

L. Jia, Z. Zhang, L. Wang, W. Jiang, M. Zhao , “Adaptive Neighborhood Propagation by Joint L2,1 -norm Regularized Sparse Coding for Representation and Classification,” in: Proceedings of IEEE International Conference on Data Mining (ICDM), Barcelona, Spain, 2016

2016
[37]

Robust Alternating Low- Rank Representation by Joint Lp - and L2,p-norm Minimization,

Z. Zhang, M. Zhao, F. Li, L. Zhang and S. Yan, "Robust Alternating Low- Rank Representation by Joint Lp - and L2,p-norm Minimization," Neural Networks (NN), vol.96, pp.55-70, Dec 2017

2017
[38]

Similarity-Adaptive Latent Low -Rank Representation for Robust Data Representation,

L. Wang, Z. Zhang, S. Li, G. Liu, C. Hou and J. Qin, "Similarity-Adaptive Latent Low -Rank Representation for Robust Data Representation," In: Proceedings of the 15th Pacific Rim International Conference on Artificial Intelligence (PRICAI), Nanjing, China, June 2018

2018
[39]

Local Low -Rank Hawkes Processes for Temporal User-Item Interactions,

J. Shang and M. Sun, “Local Low -Rank Hawkes Processes for Temporal User-Item Interactions,” in: Proceedings of the IEEE International Conference on Data Mining (ICDM), pp.427-436, 2018

2018
[40]

Robust Ne ighborhood Preserving Projection by Nuclear/L2,1 -Norm Regularization for Image Feature Extraction,

Z. Zhang, F. Li, M. Zhao, L. Zhang and S. Yan, "Robust Ne ighborhood Preserving Projection by Nuclear/L2,1 -Norm Regularization for Image Feature Extraction," IEEE Transactions on Image Processing (IEEE TIP), vol.26, no.4, pp.1607-1622, April 2017

2017
[41]

Q. Yao, J. T. Kwok, W. Zhong, "Fast Low -Rank Matrix Learning with Nonconvex Regularization, in: Proceedings of the IEEE International Conference on Data Mining (ICDM), pp.539-548, 2015

2015
[42]

Robust Auto-weighted Projective Low -Rank and Sparse Recovery for Visual Representation,

L. Wang, B. Wang, Z. Zhang, Q. Ye, L. Fu, G. Liu and M. Wang, "Robust Auto-weighted Projective Low -Rank and Sparse Recovery for Visual Representation," Neural Networks (NN), vol.117, pp.201 -215, Sep 2019

2019
[43]

Scalable Algorithms for Locally Low- Rank Matrix Modeling,

Q. Gu, J. D. Trzasko, A. Banerjee, "Scalable Algorithms for Locally Low- Rank Matrix Modeling," in: Proceedings of the IEEE International Conference on Data Mining (ICDM), pp.137-146, 2017

2017

Pith tools

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

[1] [1]

Robust principal component analysis?

E. J. Candes, X. Li, Y. Ma, and J. Wright, “Robust principal component analysis?” Journal of the ACM, vol. 58, no. 3, pp. 1-11, 2011

work page 2011

[2] [2]

Robust principal component analysis: Exact recovery of corrupted low -rank matrices via convex optimization,

J. Wright, A. Ganesh, S. Rao, Y. Peng, and Y. Ma, “Robust principal component analysis: Exact recovery of corrupted low -rank matrices via convex optimization,” in Advances in neural information processing systems, pp. 2080-2088, 2009

work page 2009

[3] [3]

Robust subspace segmentation by low -rank representation,

G. Liu, Z. Lin, and Y. Yu, “Robust subspace segmentation by low -rank representation,” in : Proceedings of the 27th international conference on machine learning (ICML-10), pp. 663-670, 2010

work page 2010

[4] [4]

Structure -constrained low -rank representation,

K. Tang, R. Liu, Z. Su, and J. Zhang, “Structure -constrained low -rank representation,” IEEE Transactions on Neural Networks and Learning Systems, vol. 25, no. 12, pp. 2167-2179, 2014

work page 2014

[5] [5]

Subspace clustering by bl ock diagonal representation,

C. Lu, J. Feng, Z. Lin, T. Mei, and S. Yan, “Subspace clustering by bl ock diagonal representation,” IEEE Trans. Pattern Anal. Mach. Intell., vol. 41, no. 2, pp. 487–501, 2019

work page 2019

[6] [6]

Inductive robust principal component analysis,

B.-K. Bao, G. Liu, C. Xu, S. Yan, “Inductive robust principal component analysis,” IEEE Transactions on Image Processing , vol. 21, no. 8, pp. 3794–3800, 2012

work page 2012

[7] [7]

Latent low -rank representation for subspace segmentation and feature extraction,

G. C. Liu and S. C. Yan, “Latent low -rank representation for subspace segmentation and feature extraction,” in: Proceedings of the International Conference on Computer Vision, pp. 1615–1622, 2011

work page 2011

[8] [8]

Subspace clustering based on latent low rank representation with Frobenius norm minimization,

S. Yu and Y. Q. Wu, “Subspace clustering based on latent low rank representation with Frobenius norm minimization,” Neurocomputing , vol. 275, pp. 2479–2489, 2018

work page 2018