Pith. sign in

REVIEW 3 major objections 5 minor 45 references

Learning Structured Twin-Incoherent Twin-Projective Latent Dictionary Pairs for Classification

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 By learning a class-specific salient-feature projection alongside the usual analysis dictionary, and forcing both coefficients and features of other classes toward zero, TP-DPL unifies feature extraction, representation, and…

desk verdict Incremental but plausible dictionary-pair model; the theory has an unaddressed scale-invariance hole, yet the empirical story is consistent enough to warrant a serious look. read the letter →

arxiv 1908.07878 v1 pith:RO2D5R2P submitted 2019-08-21 cs.CV cs.LG

Zhao Zhang , Yulin Sun , Zheng Zhang , Yang Wang , Guangcan Liu , Meng Wang This is my paper · ORCID
classification cs.CVcs.LG
keywords structuredtwin-incoherencetwin-projectivelatentdictionarypairlearningadaptiveweightingdiscriminativeclassificationimagesalientfeatureextractionreconstructionresidualclassifier
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 TP-DPL, a dictionary pair learning method that learns, for every class, an analysis dictionary, a synthesis dictionary, and a separate salient-feature projection, and couples them with twin-incoherence penalties that push both coding coefficients and salient features of other classes toward zero. The authors claim this unifies feature extraction, representation, and classification into one objective, while a flexibly relaxed reconstruction error with bias terms avoids overfitting. In testing, a query is labeled by the class whose twin-projective reconstruction residual is smallest, avoiding any extra sparse-coding step. They report higher accuracy than prior dictionary learning methods on face, object, and scene benchmarks, for example 95.0% versus 90.9% for ADDL on UMIST. If the claims hold, dictionary learning retains a simple analytic pipeline while gaining discriminative power from structured projections.

What carries the argument

The central object is the class-specific twin projection pair $(L_l, P_l)$ acting on the same data: the analysis dictionary $L_l$ produces latent codes $L_l X_l$, and the projection $P_l$ produces salient features $P_l X_l$. Twin-incoherence forces both to annihilate the complementary data $\bar{X}_l$, which makes the reconstruction residual $\|y - D_i L_i y - P_i y\|_2^2$ the classifier. The relaxed flexible errors with bias terms and the centering matrix $H_e = I - ee^T/N$ replace the direct reconstruction and the omitted atom-norm constraint, respectively; the update equations (15)-(22) carry the alternating optimization.

What would settle it

Train TP-DPL on a class-balanced subset of UMIST while recording the Frobenius norms of $D_l$, $L_l$, and $P_l$ at every iteration; if $\|L_l\|_F$ and $\|P_l\|_F$ shrink toward zero while $\|D_l\|_F$ grows without bound, the twin-incoherence penalties are being satisfied trivially and the centering matrix is not doing the stabilizing work the paper assigns to it. Alternatively, rerun the same experiment with the explicit constraint $\|d_i\|_2^2 \le 1$ restored; a material drop in accuracy would indicate the original results depended on the omitted constraint.

Watch

Extended reading notes

Core claim

TP-DPL extends projective dictionary pair learning by adding, for each class $l$, a feature projection $P_l$ alongside the analysis dictionary $L_l$ and synthesis dictionary $D_l$, and minimizes the twin-incoherence-constrained, flexibly-relaxed reconstruction error. The relaxed objective replaces the hard reconstruction $\|X_l - D_l L_l X_l\|_F^2$ with $\|X_l + a_l e^T - D_l S_l - P_l X_l\|_F^2$ and $\|L_l X_l + b_l e^T - S_l\|_F^2$, where $a_l, b_l$ are learned bias vectors, and imposes twin-incoherence penalties $\|L_l \bar{X}_l\|_F^2 + \|P_l \bar{X}_l\|_F^2$ so that other classes' data are projected near zero in both code and feature spaces. An adaptive weight matrix $W_l$ is shared between the two spaces to preserve local neighborhoods. After training, a test sample $y$ is classified by the class $i$ that minimizes $\|y - D_i L_i y - P_i y\|_2^2$. The paper argues this integrates salient feature extraction, representation, and classification, and reports improved accuracy on seven public databases.

Load-bearing premise

The central claim depends on the assertion that the centering matrix $H_e$ alone can take over the stabilizing role of the dropped dictionary-atom norm constraint, so that the learned dictionaries do not collapse to degenerate solutions while the twin-incoherence penalties are trivially satisfied.

Editorial extensions

If this is right

  • TP-DPL's class-specific residual, $\|y - D_i L_i y - P_i y\|_2^2$, classifies a new sample directly with no extra sparse reconstruction, making online prediction fast.
  • Twin-incoherence on codes and features yields embeddings with high intra-class compactness and inter-class separation, as the clustering experiments on AR and CMU PIE with convolutional features indicate.
  • Using Frobenius norms instead of $\ell_0$/ $\ell_1$ terms keeps training efficient while preserving the block-diagonal structure of the codes.
  • TP-DPL reports higher average accuracy than DPL, ADDL, LLC-DL, and LRSDL on the face, object, and scene databases evaluated.

Reading between the lines

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

  • The same twin-incoherence idea could be applied to deep networks by adding a second projection head whose outputs are pushed to annihilate other classes, yielding more separable features without changing the backbone (an extension the paper does not explore).
  • The safety of omitting the atom-norm constraint is testable by monitoring norms during training; such a check would determine whether the reported gains come from the structured projections or from an implicit regularizing effect of the centering matrix.
  • The decomposition into latent codes and salient features made by the residual classifier suggests a natural anomaly-detection signal: samples that reconstruct poorly under every class-specific pair $(D_i, L_i, P_i)$ are out-of-distribution, which the paper does not test.
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.

Referee Report

3 major / 5 minor

Summary. The paper proposes TP-DPL, an extension of projective dictionary pair learning (DPL) that adds a twin-projective latent reconstruction term, twin-incoherence penalties on both coding coefficients and salient features, and an adaptive neighborhood-preserving weighting term. The model is trained by alternating closed-form updates over D, S, P, L, and W, and classification of a test sample y is performed by the class-specific reconstruction residual argmin_i ||y - D_i L_i y - P_i y||_2^2. The authors claim that TP-DPL unifies salient feature extraction, representation, and classification, and report state-of-the-art accuracies on several face, object, and scene benchmarks.

Significance. If the model and its optimization were sound, the proposed unification of dictionary pair learning with salient-feature projection and twin-incoherence would be a reasonable contribution to the discriminative dictionary learning literature. The paper is clearly organized, the objective and update equations are specified in detail, and the experimental evaluation spans many public benchmarks with comparisons to published results. The main mathematical guarantees, however, currently rest on an unjustified removal of the atom-norm constraint and on an incorrect coercivity claim. These issues are load-bearing for the convergence proof and for the meaningfulness of the learned dictionary, so the significance of the empirical claims is conditional on a corrected formulation.

major comments (3)
  1. [§III-A, §III-B, Eq. (13), Remark 2] The removal of the DPL atom-norm constraint ||d_i||_2^2 <= 1 is not justified by the centering matrix H_e. H_e only removes column means; it imposes no bound on D, L, S, or P. Concretely, fix W_l=I and P_l=0, and choose L_l such that L_l \bar X_l=0, S_l=L_l X_l, and D_l L_l X_l H_e = X_l H_e. For every t>0, replacing (D_l,S_l,L_l) by (D_l/t, tS_l, tL_l) leaves the data-reconstruction term, the gamma coupling term, the alpha twin-incoherence terms, and the beta locality terms in Eq. (13) exactly unchanged, so the objective is constant while ||S_l||_F and ||L_l||_F tend to infinity and ||D_l||_F tends to zero. This directly contradicts the assertion in Remark 2 that f tends to infinity as the norms grow; the compactness/accumulation-point conclusion does not follow, and Remark 3's limit claim in Eq. (23) is therefore also unsupported. The learned dictionary atoms are not guaranteed to be meaningful without a norm constraint or an equivalent regularization. Please reintroduce the constraint ||d_i||_2^2 <= 1 or add a norm penalty on D (and possibly L), re-derive the updates, and redo the convergence analysis and experiments accordingly.
  2. [§III-C, Remark 1] The claim that the objective in Eq. (13) is biconvex with blocks (D,L,W) and (S,P) is not correct as stated. The beta term ||L_l X_l - L_l X_l W_l||_F^2 contains the product L_l W_l; the function (L_l,W_l) -> ||L_l X_l (I-W_l)||_F^2 is not jointly convex. For example, in the scalar case (l - lw)^2 has an indefinite Hessian. Thus the five-block alternating scheme does not fit the two-block ACS framework of [20-22] under the partition claimed in Remark 1. While exact minimization in each coordinate step guarantees non-increase of the objective values, the stronger conclusions of convergence to a stationary point and Eq. (23) are not established by the arguments given. Please provide a valid partitioning or a different convergence argument.
  3. [§V.C, Tables III-VIII] The empirical comparisons would be substantially stronger if the paper reported standard deviations for all tables and clarified whether every method was evaluated on exactly the same random train/test splits. Several baseline numbers are adopted from earlier papers, and the comparison protocol is not fully specified. Since the headline accuracy gains are large (for example, 95.0% versus 90.9% for ADDL on UMIST in Table VI), the reader needs assurance that the baseline configurations were tuned consistently and that the splits match exactly.
minor comments (5)
  1. [Fig. 1 caption] The caption refers to the 'RA-DPL framework' but the paper proposes TP-DPL; please correct the acronym.
  2. [§V.C, YaleB paragraph] The parameters are listed as 'α=0.0005, β=500 and λ=0.5'; the model has no λ parameter, so this should presumably be γ=0.5.
  3. [§V.D, clustering experiments] The convolutional features are attributed to LeNet-5 but references [35] and [38] do not appear to be the canonical LeNet-5 source; please cite the original architecture or a directly relevant reference.
  4. [§V.A, convergence experiment] The phrase 'averaged results over 20 iterations' is ambiguous: convergence plots normally show the objective value per iteration, and it should be stated whether the curves are averaged over multiple random initializations or over runs.
  5. [Equations throughout] Several equations contain OCR-style garbles, such as repeated H_e factors in Eq. (15) and missing minimization arguments in Eqs. (7), (8), and (13); please proofread the mathematical notation carefully.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation found: TP-DPL's residual classifier is trained on labeled data and evaluated on held-out test samples.

full rationale

The claimed prediction chain is not circular. Eqs. (7)/(13) define a supervised objective: for each class l, the model fits D_l, L_l, P_l, S_l, and W_l to reconstruct X_l while penalizing the response of other classes' data. The classifier Eq. (24), identity(y)=argmin_i ||y-D_i L_i y-P_i y||_2^2, applies the trained class-specific dictionaries and projections to a test sample y whose label was never used in the optimization; the residual is a genuine out-of-sample quantity rather than a fitted input renamed as a prediction. The method's components (twin-incoherence, adaptive graph weights, flexible bias terms) are added to the objective before classification, and test accuracy is an external empirical outcome, not an identity. The references to the authors' own prior work (ADDL [11] and LLC-DL [10]) appear as comparison baselines and modeling context, not as a load-bearing uniqueness theorem or ansatz that forces Eq. (24); the baseline numbers are adopted from public benchmark papers, so the state-of-the-art claim is falsifiable outside the model's own fitted values. The objection that omitting DPL's atom-norm constraint ||d_i||_2^2<=1 in Sec. III-A and replacing it with the centering matrix H_e in Sec. III-B may break coercivity is a legitimate optimization and correctness concern, but it is not circularity: even if Remark 2's compactness argument fails, the classifier is not defined in terms of the quantity it is supposed to predict. No circular step satisfying the quoted-reduction standard was found.

Assumptions & free parameters 4 free parameters · 4 assumptions · 2 invented entities

The model depends on several tuned hyperparameters (alpha, beta, gamma, K), each varying strongly across datasets. It also relies on external convergence theorems and on the unproven assumption that centering can replace DPL's atom norm constraint. The introduced P_l and W_l are learned model components without independent empirical handles.

free parameters (4)
  • alpha (twin-incoherence weight) = YaleB 0.0005; AR 0.0005; CMU PIE 50; UMIST 0.0005; Fifteen 5e-5; ETH80 50; clustering AR 50; CMU PIE 50
    Tuned per dataset by grid search in Section V-B; no validation split is described, so the tuning procedure uses test accuracy.
  • beta (neighborhood preservation weight) = YaleB 500; AR 50000; CMU PIE 500; UMIST 50; Fifteen 5000; ETH80 50; clustering 50
    Selected by the same grid search; values vary by orders of magnitude across datasets, indicating strong per-dataset tuning.
  • gamma (flexible relaxation weight) = YaleB 0.5 (noted as lambda); AR 0.5; CMU PIE 0.005; UMIST 0.0005; Fifteen 0.5; ETH80 0.05; clustering AR 0.5; CMU PIE…
    Tuned with grid search; the paper sometimes writes lambda and sometimes gamma for this weight, adding to ambiguity.
  • dictionary size K = e.g., 570 for YaleB, 500 for AR, training-set size for CMU PIE and UMIST, 450 for Fifteen, 480 for ETH80
    Chosen per dataset, typically proportional to the number of training samples; affects capacity and is not derived from data.
assumptions (4)
  • standard math Gorski et al. ACS convergence theorems (Theorems 1 and 2) apply to the TP-DPL objective.
    Section III-C uses these external theorems to claim monotonic convergence, existence of accumulation points, and local optimality.
  • domain assumption The centering matrix H_e plays the same stabilizing role as the omitted atom norm constraint ||d_i||^2 <= 1.
    Section III-A omits the DPL constraint and Section III-B claims H_e is sufficient, but no proof of boundedness or non-degeneracy is given.
  • domain assumption A single shared weight matrix W_l can reconstruct neighborhoods in both code space L_l X_l and feature space P_l X_l.
    Eq. (6) assumes one W_l per class works for both projections; no justification is provided beyond a locality-preservation heuristic.
  • domain assumption The class-specific reconstruction residual in Eq. (24) is a valid classifier for new samples.
    Assumes trained D_i, L_i, P_i generalize so that in-class residuals are smaller than out-of-class residuals for unseen data.
invented entities (2)
  • class-specific salient feature projection P_l
    purpose: Extracts a per-class latent feature subspace P_l X_l to improve reconstruction and add discriminative information.
    A learned n x n matrix per class with no external falsifiable handle; its only evidence is in-sample reconstruction and benchmark accuracy.
  • adaptive reconstruction weight matrix W_l
    purpose: Preserves local neighborhood structure shared between coefficients and salient features.
    Learned per class from training data; no independent source of validation is provided.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Learning Structured Twin-Incoherent Twin-Projective Latent Dictionary Pairs for Classification." pith.science (2026). https://pith.science/paper/RO2D5R2P

@misc{pith2026190807878,
  author       = {Pith},
  title        = {Pith review of: Learning Structured Twin-Incoherent Twin-Projective Latent Dictionary Pairs for Classification},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/RO2D5R2P}},
  note         = {Machine review of arXiv:1908.07878}
}
read the original abstract

In this paper, we extend the popular dictionary pair learning (DPL) into the scenario of twin-projective latent flexible DPL under a structured twin-incoherence. Technically, a novel framework called Twin-Projective Latent Flexible DPL (TP-DPL) is proposed, which minimizes the twin-incoherence constrained flexibly-relaxed reconstruction error to avoid the possible over-fitting issue and produce accurate reconstruction. In this setting, our TP-DPL integrates the twin-incoherence based latent flexible DPL and the joint embedding of codes as well as salient features by twin-projection into a unified model in an adaptive neighborhood-preserving manner. As a result, TP-DPL unifies the salient feature extraction, representation and classification. The twin-incoherence constraint on codes and features can explicitly ensure high intra-class compactness and inter-class separation over them. TP-DPL also integrates the adaptive weighting to preserve the local neighborhood of the coefficients and salient features within each class explicitly. For efficiency, TP-DPL uses Frobenius-norm and abandons the costly l0/l1-norm for group sparse representation. Another byproduct is that TP-DPL can directly apply the class-specific twin-projective reconstruction residual to compute the label of data. Extensive results on public databases show that TP-DPL can deliver the state-of-the-art performance.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

45 extracted references · 44 canonical work pages

  1. [1]

    Label consistent K-SVD: Learning a discriminative diction ary for recognition ,

    Z. Jiang, L. Z. Lin, and S. Davis, “Label consistent K-SVD: Learning a discriminative diction ary for recognition ,” IEEE Trans. Pattern Anal. Mach. Intell., vol. 35, no. 35, pp. 2651-2664, 2013

    work page 2013

  2. [2]

    Sparse Codes Auto -Extractor for Classification: A Joint Embedding and Dictionary Learning Frame work for Representation,

    Z. Zhang, F.Z. Li, T. W. S. Chow, L. Zhang, and S.C. Yan, “Sparse Codes Auto -Extractor for Classification: A Joint Embedding and Dictionary Learning Frame work for Representation,” IEEE Trans. Signal Processing, vol. 64, no.14, pp. 3790-3805, 2016

    work page 2016

  3. [3]

    K -SVD: An Algorithm for Designing Overcomplete Dictionaries for Sparse Representation,

    M. Aharon, M. Elad, and A. Bruckstein, “K -SVD: An Algorithm for Designing Overcomplete Dictionaries for Sparse Representation,” IEEE Trans. Signal Processing, vol. 54, no. 1, pp. 4311-4322, 2006

    work page 2006

  4. [4]

    Robust Face Recognition via Sparse Representation,

    J. Wright, M. Yang, A. Ganesh, S. Sastry, and Y. Ma, “Robust Face Recognition via Sparse Representation,” IEEE Trans. on Pattern Analysis and Machine Intelligence, vol. 31, no. 2, pp. 210-227, 2009

    work page 2009

  5. [5]

    A Locality-Constrained and Label Embedding Dictionary Learning Algorithm for Image Classification,

    Z. Li, Z. Lai, Y. Xu, J. Yang, and D. Zhang, “A Locality-Constrained and Label Embedding Dictionary Learning Algorithm for Image Classification,” IEEE Trans. Neural Networks and Learning Systems, vol.28, no.2, pp.278-293, 2017

    work page 2017

  6. [6]

    Discriminative K -SVD for Dictionary Learning in Face Recognition,

    Q. Zhang and B. Li, “Discriminative K -SVD for Dictionary Learning in Face Recognition,” In: Proc. IEEE Conf. Computer Vision and Pattern Recognition, San Francisco, CA USA, pp.2691-2698, 2010

    work page 2010

  7. [7]

    Sparse Representation based Fisher Discrimination Dictionary Learn ing for Image Classification,

    M. Zhang, L. Zhang, X. Feng, and D. Zhang, “Sparse Representation based Fisher Discrimination Dictionary Learn ing for Image Classification,” International Journal of Computer Vision , vol.109, no.3, pp. 209-232, 2014

    work page 2014

  8. [8]

    Classification and Clustering via dictionary learning with structured incoherence and shared features,

    I. Ramirez, P. Sprechmann, G. Sapiro, “Classification and Clustering via dictionary learning with structured incoherence and shared features,” In: Proc. IEEE CVPR, CA USA, pp. 3501-3508, 2010

    work page 2010

Show all 45 references
  1. [9]

    Projective dictionary pair learning for pattern classification,

    S. Gu, L. Zhang, W. Zuo, and X. Feng, “Projective dictionary pair learning for pattern classification,” In: Proceedings of the Neural Information Processing Systems, Montreal, QC, Canada, 2014

    2014

  2. [10]

    Structured Latent Label Consistent Dictionary Learning for Salient Machine Faults Representation based Robust Classification ,

    Z. Zhang, W. Jiang, F. Li, M. Zhao, B. Li and L. Zhang, “Structured Latent Label Consistent Dictionary Learning for Salient Machine Faults Representation based Robust Classification ,” IEEE Trans . on Industrial Informatics, vol.13, iss.2, pp.642-654, 2017

    2017

  3. [11]

    Jointly learning structured analysis discriminative dictionary and analysis multiclass classifier

    Z. Zhang, W. Jiang, J. Qin, L. Zhang, F. Li, M. Zhang, and S. Yan, "Jointly learning structured analysis discriminative dictionary and analysis multiclass classifier", IEEE Trans. on Neural Networks and Learning Systems, vol.29, iss.8, pp.3798-3814, 2018

    2018

  4. [12]

    Joint label consistent dictionary learning and adaptive label prediction for semi-supervised machine fault classification,

    W. Jiang, Z. Zhang, F. Li, L. Zhang, M. Zhao, and X. Jin, “Joint label consistent dictionary learning and adaptive label prediction for semi-supervised machine fault classification,” IEEE Transactions on Industrial Informatics, vol. 12, no. 1, pp. 248-256, 2016

    2016

  5. [13]

    Sparse Dictionary -Based Representation and Recognition of Action Attributes,

    Q. Qiu, Z. L. Jiang, and R. Chellappa, “Sparse Dictionary -Based Representation and Recognition of Action Attributes,” In: Proc. IEEE Int’l Conf. Computer Vision, 2011

    2011

  6. [14]

    Image denoising via sparse and redu ndant representations over learned dictionaries,

    M. Elad and M. Aharon, “Image denoising via sparse and redu ndant representations over learned dictionaries,” IEEE Transactions on Image processing, vol. 54, no. 12, pp. 3736-3745, 2006

    2006

  7. [15]

    Dictionaries for sparse representation modeling,

    R.Rubinstein, A. Bruckstein, and M. Elad, “Dictionaries for sparse representation modeling,” Proceedings of the IEEE , vol.98, no .6, pp.1045-1057, 2010

    2010

  8. [16]

    Fast low -rank shared dictionary learning for image classification,

    T. H. Vu, and V. Monga, “Fast low -rank shared dictionary learning for image classification,” IEEE Trans. on Image Processing, vol. 26, no. 11, pp. 5160-5175, 2017

    2017

  9. [17]

    Discriminative Sparse Flexible Manifold Embedding with Novel Graph for Robust Visual Representation and Label Propagation,

    Z. Zhang, Y. Zhang, F. Z. Li, M. B. Zhao, L. Zhang and S. C. Yan, “Discriminative Sparse Flexible Manifold Embedding with Novel Graph for Robust Visual Representation and Label Propagation,” Pattern Recognition, vol. 61, pp. 492–510, Jan 2017

    2017

  10. [18]

    Robust Spar se Linear Discriminant Analysis,

    J. Wen, X. Fang, J. Cui, L. Fei, K. Yan, K. Yan, Y. Chen, Y. Xu ， “Robust Spar se Linear Discriminant Analysis,” IEEE Transactions on Circuits and Systems for Video Technology, 2018

    2018

  11. [19]

    Submodular Dictionary Learning for Sparse Coding,

    Z. Jiang, G. Zhang, and L. Davis, “Submodular Dictionary Learning for Sparse Coding,” In: Proc. IEEE CVPR, 2012

    2012

  12. [20]

    Block-relaxation algorithms in statistics,

    J. Leeuw, “Block-relaxation algorithms in statistics,” In: Information systems and data analysis, Springer, pp. 308–324, 1994

    1994

  13. [21]

    Minimization of a non -separable objective function su bject to disjoint constraints,

    R.E. Wendell, and A.P. Hurter Jr, “Minimization of a non -separable objective function su bject to disjoint constraints,” Operations Research, vol.24, no. 4, pp. 643-657, 1976

    1976

  14. [22]

    Biconvex sets and optimization with biconvex functions: a survey and extensions,

    J. Gorski, F. Pfeuffer, K. Klamroth, “Biconvex sets and optimization with biconvex functions: a survey and extensions,” Mathematical Methods of Operations Research. Vol. 66, pp. 373-407, 2007

    2007

  15. [23]

    Document clustering using locality preserving indexing,

    D. Cai, X. He,J. Han, “Document clustering using locality preserving indexing,” IEEE Transactions on Knowledge and Data Engineering , vol.17, no.12, pp.1624-1637, 2005

    2005

  16. [24]

    Laplacian score for feature selection,

    X. He, D. Cai, P. Niyogi, “Laplacian score for feature selection,” In: Advances in neural information processing systems, 2005

    2005

  17. [25]

    Parameterisation of a stochastic model for human face identification,

    F. Samaria, and A. Harter, “Parameterisation of a stochastic model for human face identification,” In: Proc. IEEE Winter Conf. on Applications of Computer Vision, pp.138-142, 1994

    1994

  18. [26]

    From Few to Many: Illumination cone Models for Face Recognication under Variable Lighting and Pose,

    A.Georghiades, P. Belhumeur and D. Kriegman, “From Few to Many: Illumination cone Models for Face Recognication under Variable Lighting and Pose,” IEEE Trans. on Pattern Analysis and Machine Intelligence, vol.23, no.6, pp.643-660, 2001

    2001

  19. [27]

    Characterizing virtual eigensign atures for general purpose face recognition,

    D. Graham, and N. Allinson, “Characterizing virtual eigensign atures for general purpose face recognition,” Face Recognition: From Theory to Applications, volume 163 of NATO ASI Series F, Computer and Systems Sciences, pp.446-456, 1998

    1998

  20. [28]

    The AR face database,

    A. Martinez, “The AR face database,” CVC Technical Report, 1998

    1998

  21. [29]

    The CMU pose, illumination, and expression (PIE) database,

    T. Sim, S. Bak er, and M. Bsat, “The CMU pose, illumination, and expression (PIE) database,” In: Proceedings of IEEE International Conf. on Automatic Face and Gesture Recognition, pp.53-58, 2002

    2002

  22. [30]

    Analyzing appearance and contour based methods f or object categorization,

    B. Leibe and B. Schiele, “Analyzing appearance and contour based methods f or object categorization,” In: Proc. IEEE Conf. Computer Vision and Pattern Recognition, Madison, WI, USA, 2003

    2003

  23. [31]

    Beyond Bags of Features: Spatial Pyramid Matching for Recognizing Natural Scene Categories,

    S. Lazebnik, C. Schmid, J. Ponce, “Beyond Bags of Features: Spatial Pyramid Matching for Recognizing Natural Scene Categories,” In: Proc. IEEE Conf. CVPR, New York, NY, pp.2169-2178, 2006

    2006

  24. [32]

    A dictionary learning approach for classification separating the particularity and the commonality,

    S. Kong, D. Wang, “A dictionary learning approach for classification separating the particularity and the commonality,” In: Proc. European Conference on Computer Vision, Springer, pp. 186-199, 2012

    2012

  25. [33]

    Introduction to Statistical Pattern Recognition ,

    K. Fukunaga, “Introduction to Statistical Pattern Recognition ,” 2nd ed., Boston, MA, USA: Academic, 2002

    2002

  26. [34]

    Uncertainty principles and ideal atomic decomposition

    D. L. Donoho and X. Huo, "Uncertainty principles and ideal atomic decomposition", IEEE Trans. on Information Theory, vol. 47, no. 7 , pp. 2845-2862, Nov. 2001

    2001

  27. [35]

    Reconstructible Nonlinear Dimensionality Reduction via Joint Dictionary Learning,

    X. Wei, H. Shen, Y. Li, X. Tang, F. Wang, M. Kleinsteuber and Y. Lu Murphey, “Reconstructible Nonlinear Dimensionality Reduction via Joint Dictionary Learning,” IEEE Trans. Neural Netw. Learning Syst., vol.30, no.1, pp.175-189, 2019

    2019

  28. [36]

    Robust Locality-Constrained Label Consistent KSVD by Joint Sparse Embedding,

    Z. Zhang, W. Jiang, S. Li, J. Qin, G. C. Liu and S. C. Yan, "Robust Locality-Constrained Label Consistent KSVD by Joint Sparse Embedding," In: Proceedings of ICPR, Beijing, China, 2018

    2018

  29. [37]

    Comp onent-based Face Recognition with 3D Morphable Models,

    B. Weyrauch, J. Huang, B.Heisele, and V. Blanz, “Comp onent-based Face Recognition with 3D Morphable Models,” In: Proceedings of the IEEE Workshop on Face Processing in Video, Washington, 2004

    2004

  30. [38]

    Discriminative Robust Deep Dictionary Learning for Hyperspectra l Image Classification,

    V. Singhal, K. Aggarwal H, S. Tariyal, M. Angshul, “Discriminative Robust Deep Dictionary Learning for Hyperspectra l Image Classification,” IEEE Trans. on Geoscience and Remote Sensing, vol. 55, no. 9, pp. 5274-5283, 2017

    2017

  31. [39]

    Contextual Online Dictionary Learning for Hyperspectral Image Classification,

    W. Fu, S. Li, L. Fang, and J. Atli Benediktsson, “Contextual Online Dictionary Learning for Hyperspectral Image Classification,” IEEE Transactions on Geoscience and Remote Sensing, vol. 56, no. 3, pp. 1336-1347, 2018

    2018

  32. [40]

    Imagenet classification with deep convolutional neural networks,

    A. Krizhevsky, I. Sutskever and G. Hinton, “Imagenet classification with deep convolutional neural networks,” In: Advances in Neural Information Processing Systems, pp. 1106-1114, 2012

    2012

  33. [41]

    Robust Projective Low-Rank and Sparse Representation by Robust Dictionary Learning,

    J. Ren, Z. Zhang, S. Li, G. Liu, M. Wang, S. Yan, "Robust Projective Low-Rank and Sparse Representation by Robust Dictionary Learning," In: Proceedings of ICPR, Beijing, China, 2018

    2018

  34. [42]

    Self -Expressive Dictionary Learning for Dynamic 3D Reconstruction,

    E. Zheng, D. Ji, E. Dunn, and J. Frahm, “Self -Expressive Dictionary Learning for Dynamic 3D Reconstruction,” IEEE Trans. Pattern Anal. Mach. Intell., vol.40, no.9, pp.2223-2237, 2018

    2018

  35. [43]

    Joint Subspace Recovery and Enhanced Locality Driven Robust Flexible Discriminative Dictionary Learning,

    Z. Zhang, J. Ren, W. Jiang, Z. Zhang, R. Hong, S. Yan and M. Wang, "Joint Subspace Recovery and Enhanced Locality Driven Robust Flexible Discriminative Dictionary Learning," IEEE Transactions on Circuits and Systems for Video Technology, June 2019

    2019

  36. [44]

    Scalable Block- Diagonal Locality -Constrained Projective Dictionary Learning,

    Z. Zhang, W. Jiang, Z. Zhang, S. Li, G. Liu, J. Qin, "Scalable Block- Diagonal Locality -Constrained Projective Dictionary Learning," In: Proceedings of the 28th Intern ational Joint Conference on Artificial Intelligence (IJCAI), Macao, China, Aug 2019

    2019

  37. [45]

    Discriminative Fisher Embedding Dictionary Learning Algorithm for Object Recognition,

    Z. Li, Z. Zhang, J. Qin, Z. Zhang and L. Shao, "Discriminative Fisher Embedding Dictionary Learning Algorithm for Object Recognition," IEEE Transactions on Neural Networks and Learning Systems (IEEE TNNLS), April 2019. DOI: 10.1109/TNNLS.2019.2910146

    2019

Pith tools

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