Pith. sign in

REVIEW 2 major objections 5 minor 28 references

Efficient iterative techniques for solving tensor problems with the T-product

T0 review · 2 major / 5 minor · reviewed 2026-08-16 · deepseek-v4-flash

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

Pith's one-line read For $C\star X=D$ under the T-product, tensor-only iterative updates reach an exact solution in at most $n_1\ell n_3+1$ steps, and least-squares problems in at most $n_2\ell n_3$ steps.

desk verdict A competent but unsurprising translation of CG/CGNE/CGLS to the T-product tensor setting, with standard exact-arithmetic proofs, overshadowed by an overclaimed image-restoration experiment. read the letter →

arxiv 2504.17861 v1 pith:YVTXOPA3 submitted 2025-04-24 math.NA cs.NA

Malihe Nobakht Kooshkghazi , Salman Ahmadi-Asl , Hamidreza Afshin This is my paper · ORCID
classification math.NAcs.NA MSC 15A6965F10
keywords T-producttensorequationiterativealgorithmleast-squaressolutionminimalFrobeniusnormorthogonalsequencefiniteterminationimagerestoration
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

This paper is about solving tensor equations of the form $C\star X=D$, where $\star$ is the T-product, a tensor multiplication defined by block-circulant convolution along the third dimension. The authors propose iterative algorithms that work directly on tensors, never forming the large block-circulant matrix, and they claim each algorithm stops in a provably finite number of iterations in exact arithmetic: at most $n_1\ell n_3+1$ steps for a consistent equation, and at most $n_2\ell n_3$ steps for the least-squares problem $\min_X \|C\star X-D\|_F$. The practical payoff would be a memory-lean, predictable solver for tensor systems arising in image and video restoration, where starting from the zero tensor automatically selects the minimal Frobenius norm solution. The arguments turn on orthogonality: the generated residuals are mutually orthogonal, and the search directions are mutually orthogonal under the T-product inner product, so the residual sequence cannot continue beyond the dimension of the tensor space.

What carries the argument

The load-bearing object is the T-product together with its Fourier-diagonalization identity: after applying fft along the third dimension, $C\star D$ is computed slicewise as $\widehat C^{(i)}\widehat D^{(i)}$, so iterations can be coded directly on tensors. The algorithmic core is the direction-update recurrence $X_k=X_{k-1}+\frac{\|R_{k-1}\|^2}{\|Q_{k-1}\|^2}Q_{k-1}$, with $R_k=D-C\star X_k$, $P_k=C^T\star R_k$, and $Q_k=P_k-\frac{\langle P_k,Q_{k-1}\rangle}{\|Q_{k-1}\|^2}Q_{k-1}$. The inner-product identity $\langle C,D\rangle=\operatorname{Tr}((D^T\star C)^{(1)})$ translates every orthogonality statement into a trace condition, allowing the proofs to run in tensor arithmetic; the same recurrence, applied to the normal equation, gives the least-squares variant.

What would settle it

Run Algorithm 3 in exact rational arithmetic on a small generic consistent equation, for example $n_1=n_2=n_3=2$, $\ell=1$, with $D$ constructed as $C\star X^*$ for a known $X^*$, and check whether the residual after $n_1\ell n_3+1=5$ steps is exactly zero for a nonzero starting tensor; any nonzero exact residual would disprove the finite-termination theorem. For the least-squares claim, run Algorithm 4 with $X_1=O$ in the same exact setting and verify by an independent calculation that the returned tensor minimizes $\|C\star X-D\|_F$ and is the unique minimal-norm minimizer.

Watch

Extended reading notes

Core claim

The central claim is that direction-update tensor recurrences terminate exactly: for a consistent equation $C\star X=D$ with $C\in\mathbb{R}^{n_1\times n_2\times n_3}$, Algorithm 3 returns an exact solution in at most $n_1\ell n_3+1$ steps from any initial tensor, and for the possibly inconsistent case Algorithm 4 returns a least-squares solution of $\min_X\|C\star X-D\|_F$ in at most $n_2\ell n_3$ steps. The mechanism is a dimension argument: the residual tensors $R_k$ are mutually orthogonal in the T-product inner product, and the search directions $Q_k$ are likewise orthogonal, so within a space of dimension $n_1\ell n_3$ one residual must vanish; in the least-squares case the same argument runs on the normal equation $C^T\star C\star X=C^T\star D$. Choosing $X_1=C^T\star H$, in particular $X_1=O$, forces the returned solution into the range of $C^T$, which the paper proves is exactly the minimal Frobenius norm solution, and a residual with $Q_k=O$ signals inconsistency.

Load-bearing premise

The iteration counts and minimal-norm conclusions assume exact arithmetic, because the proofs require the residual and search-direction tensors to be exactly orthogonal; with floating-point arithmetic those orthogonalities are lost to rounding, so the stated guarantees are only approximate and the paper gives no rounding-error analysis.

Editorial extensions

If this is right

  • Consistent T-product systems inherit a hard iteration bound: from any starting tensor the residual is exactly zero after at most $n_1\ell n_3+1$ updates, so a solvable system cannot make the solver loop indefinitely.
  • Because the algorithms never form the equivalent block-circulant or Kronecker matrices, per-iteration storage stays at the size of the tensors themselves, and the dominant cost is a handful of fft-based T-products per step.
  • Starting from the zero tensor automatically returns the minimal Frobenius norm solution of a consistent equation and the minimal-norm least-squares solution of an inconsistent one, giving a deterministic route to a pseudoinverse-type tensor solution.
  • Algorithm 4 also covers the consistent case, since any exact solution minimizes the residual; a single implementation therefore handles both solvable and unsolvable systems, with the condition $R_k\neq O$, $Q_k=O$ marking inconsistency.

Reading between the lines

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

  • Beyond the paper, the termination bound is essentially a dimension count, so the same recurrence should transfer to other tensor products whenever a compatible transpose and trace inner product exist; the iteration limit would be the dimension of the corresponding solution space.
  • Beyond the paper, the reported experiments use random Gaussian coefficient tensors, which are typically well-conditioned; the natural stress test not included is to run Algorithm 4 on severely ill-conditioned or structured coefficient tensors and measure how far the floating-point residual deviates from the exact-arithmetic bound.
  • Beyond the paper, the minimal-norm guarantee suggests a direct method for the T-product pseudoinverse: apply Algorithm 4 with $X_1=O$ to $C\star X=I$ for multiple right-hand sides and assemble the returned slices; the same finite-termination bound would apply.
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

2 major / 5 minor

Summary. The paper develops three iterative algorithms for tensor equations under the T-product. Algorithm 2 is a conjugate-gradient-type method for C*X=D when C is T-symmetric positive definite; Algorithm 3 treats consistent equations C*X=D with rectangular C by a CGNR-type process using residuals R_k=D-C*X_k and search directions generated from C^T*R_k; Algorithm 4 applies the same construction to the normal equation C^T*C*X=C^T*D to obtain least-squares solutions. The main formal claims are that Algorithm 3 terminates in at most n1*l*n3+1 iterations with the exact solution, that Algorithm 4 terminates in at most n2*l*n3 iterations with a least-squares solution, and that with X1=0 the computed solutions have minimal Frobenius norm. Numerical experiments on random synthetic tensors and on image/video recovery illustrate the residual decay.

Significance. Read as exact-arithmetic statements, the convergence and minimal-norm results are coherent, and the tensor formulation is convenient because the algorithms avoid explicit construction of bcirc matrices and can be implemented with the FFT-based T-product. The extension of classical CG/CGNR to the T-product is natural, the proofs are self-contained, and the paper gives explicit finite-termination bounds that are useful for this problem class. A manual consistency check of the recurrences on a small system indicates that Algorithm 3 behaves as stated in exact arithmetic. The novelty is incremental, and no reproducibility artifact or comparison with existing T-product solvers is provided, but the paper's central exact-arithmetic claims are defensible. The main substantive weakness is that the practical claims of finite termination with negligible errors are stated without any treatment of floating-point effects, so the theorems should be read as exact-arithmetic results rather than as unconditional statements about the implemented algorithms.

major comments (2)
  1. [Theorems 4.4 and 5.4; Algorithms 3 and 4; Section 6] The termination and minimal-norm guarantees are exact-arithmetic statements. The recurrences in Algorithm 3, step 5, and Algorithm 4, step 5, are used to prove <R_i,R_j>=0 and <Q_i,Q_j>=0 in Lemma 4.3; in floating-point arithmetic these identities are destroyed by rounding in the T-product, the inner products, and the short recurrences, and no reorthogonalization or rounding-error analysis is supplied. Consequently the asserted iteration bounds (n1*l*n3+1 and n2*l*n3) and the minimal-norm conclusion that relies on exact termination at a zero residual are not guaranteed by the implemented algorithms. Because the abstract and Section 6 present the methods as practical solvers with 'negligible errors,' this gap is load-bearing; either add a stability or rounding-error discussion, or explicitly qualify the theorems as exact-arithmetic results and soften the practical claim.
  2. [Theorem 4.6, Eqs. (4.6)-(4.7)] The proof of minimal Frobenius norm relies on the equivalence C*X=D <=> C.X=D <=> (I_{l.n3} (x) C) vec(X) = vec(D). For the T-product, the correct linear representation is bcirc(C) unfold(X) = unfold(D) in the original domain, or bdiag(C_hat) X_hat = D_hat in the Fourier domain; it is not simply (I_{l.n3} (x) C) vec(X) as written, unless substantial additional notation is introduced. As written, the range argument for the adjoint is therefore not established. The result itself is true: starting from X1=0, all iterates lie in the range of Y -> C^T*Y, and the solution in that range has minimal Frobenius norm. Please replace the proof with a correct argument.
minor comments (5)
  1. [Abstract and throughout] The abstract contains the typo 'demonste' for 'demonstrate,' and the text has numerous other typos, including 'tensot,' 'researchears,' 'uder,' and 'edequate'; the manuscript should be carefully proofread.
  2. [Lemma 3.3] In the proof, the displayed expression for Tr[(f(Y)-f(X))^{(1)}] has an extra factor 1/2 in front of the first term; the following line drops the factor, so the conclusion is correct but the displayed equation should be fixed.
  3. [Theorem 3.4] In the proof, the expression 'Tr((P_{k+1} * A * P_k)(1)) = 0' uses an undefined tensor A; it should presumably be the C-orthogonality statement Tr((P_{k+1}^T * C * P_k)(1)) = 0.
  4. [Section 3] Algorithm 2 is not given an explicit finite-termination theorem, although the same dimension argument used in Theorem 4.4 applies; stating this theorem would make the abstract's 'finite number of iterations' claim complete for all three algorithms.
  5. [Theorem 5.4] The proof is only one sentence referring to Theorems 4.4 and 4.6; because the normal-equations operator C^T*C can be semidefinite in the rank-deficient case, the proof should include a short explanation of why the dimension count and the minimal-norm argument remain valid.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the algorithms' convergence and termination proofs are self-contained, and no fitted quantities are relabeled as predictions.

full rationale

The paper's derivation chain is self-contained. The T-product properties used as inputs (distributivity, associativity, transpose, inner-product identity, DFT block-diagonal equivalence) are cited standard results from the literature, not derived from the paper's own conclusions. Algorithms 2–4 are constructed by explicitly deriving residual recurrences and scalar step/orthogonalization coefficients from the definitions, and the finite-termination theorems (Theorem 3.4, Lemma 4.3, Theorem 4.4, Theorem 5.4) prove orthogonality of the generated residual and search-direction sequences by induction inside the paper, then invoke only the finite dimensionality of the tensor space to conclude termination. There is no fitted parameter that is later called a prediction, no quantity defined in terms of the claimed output, and no load-bearing self-citation: the authors' own prior work [20] is cited only as context for existing Krylov-type methods for T-product equations, and the central proofs do not depend on it. The main caveat, that the exact-arithmetic orthogonality arguments do not automatically survive floating-point roundoff without reorthogonalization or a rounding-error analysis, is a numerical robustness issue rather than a circularity issue; it does not make the derivation equivalent to its inputs. Numerical experiments in Section 6 provide external demonstrations of practical behavior, but they are not needed to make the mathematical derivation circular. Overall, no circular step satisfying the required evidence standard is present.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

The central claim rests on standard T-product algebra and exact arithmetic. No free parameters are fitted; the algorithms use a user-supplied tolerance 'tol' in the stopping rule, which does not affect the convergence guarantees. Axioms are the standard definitions of T-product, DFT properties, and the consistency assumptions for Algorithms 3 and 4.

assumptions (5)
  • standard math The T-product of third-order tensors is equivalent to block-circulant matrix multiplication after unfolding, and via fft it is equivalent to slice-wise matrix multiplication of the block-diagonal Fourier representation (Definitions 2.2, 2.3, and Algorithm 1).
    This is the standard definition of the T-product from [12, 13]; the paper relies on it throughout.
  • standard math For real tensors, the Frobenius inner product satisfies <C,D> = (1/n3) <C,D> with the Fourier block-diagonal representations (Definition 2.9).
    This is a known property of the DFT; used in Corollary 2.13 and in the step-size derivations.
  • standard math The trace identity Tr((C⋆D)(1)) = <C,D> (Remark 2.10) holds.
    Used throughout the proofs to convert between trace-based expressions and inner products.
  • domain assumption The sequences generated by the algorithms are assumed to be computed in exact arithmetic; floating-point rounding is not analyzed.
    The finite-termination theorems (Theorem 3.4, 4.4, 5.4) rely on exact orthogonality, which holds only in exact arithmetic.
  • domain assumption For Algorithm 3, the equation C⋆X = D is assumed consistent (Section 4); for Algorithm 4, the normal equation is consistent by construction.
    The finite-termination guarantee for Algorithm 3 explicitly assumes consistency; inconsistency is detected by Q_k = O with R_k ≠ O (Corollary 4.5).

how reviews work

0 comments
Cite this review

Pith. "Pith review of Efficient iterative techniques for solving tensor problems with the T-product." pith.science (2026). https://pith.science/paper/YVTXOPA3

@misc{pith2026250417861,
  author       = {Pith},
  title        = {Pith review of: Efficient iterative techniques for solving tensor problems with the T-product},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/YVTXOPA3}},
  note         = {Machine review of arXiv:2504.17861}
}
read the original abstract

This paper presents iterative methods for solving tensor equations involving the T-product. The proposed approaches apply tensor computations without matrix construction. For each initial tensor, these algorithms solve related problems in a finite number of iterations, with negligible errors. The theoretical analysis is validated by numerical examples that demonstrate the practicality and effectiveness of these algorithms.

Figures

Figures reproduced from arXiv: 2504.17861 by the authors.

Figure 1
Figure 1. The norm of residual for example 1. Example 6.1. Assume that n1 = 5, n2 = 4, n3 = 3, l = 5. We randomly constructed a tensor C by the MATLAB package randn.m: C(:, :, 1) =       1.7380 −10.6399 1.4411 0.4655 −0.9092 −5.8846 −7.9709 −1.8908 −4.6977 −4.9527 0.5511 −7.4134 −0.1877 −5.8652 3.9353 −0.2191 −9.4815 −8.6271 −0.0111 4.8041       C(:, :, 2) =       8.6912 −1.1348 −6.6081 3.8850 −2.1510 −5.744… view at source ↗
Figure 2
Figure 2. The relative error for example 2. As shown in [PITH_FULL_IMAGE:figures/full_fig_p017_2.png] view at source ↗
Figure 3
Figure 3. Benchmark images used in our simulations [PITH_FULL_IMAGE:figures/full_fig_p018_3.png] view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: The original, the degraded and the recovered images using the proposed iterative method [PITH_FULL_IMAGE:figures/full_fig_p018_4.png]
Figure 5
Figure 5. Figure 5: The relative error history of the proposed algorithm for four benchmark images [PITH_FULL_IMAGE:figures/full_fig_p019_5.png]
Figure 6
Figure 6. Figure 6: Some sample frames of “Foreman” (top figure) and ”News” (bottom figure) videos [PITH_FULL_IMAGE:figures/full_fig_p019_6.png]
Figure 7
Figure 7. Figure 7: The original, the degraded and the recovered frame number 1 (as one example) using the proposed iterative method. All recovered frames had the same quality as the frame number one [PITH_FULL_IMAGE:figures/full_fig_p020_7.png]
Figure 8
Figure 8. Figure 8: The relative error history of the proposed algorithm for two benchmark videos. 7. Conclusion In this paper, we presented iterative algorithms for solving tensor equations via the T-product. For each initial tensor, these algorithms provided a sequence Xk converging to …

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

28 extracted references · 26 canonical work pages

  1. [20]

    Computational and Applied Mathematics, 42(8), 358 (2023)

    Nobakht-Kooshkghazi, M., Afshin, H., The new Krylov subspace methods for solving tensor equations via T- product. Computational and Applied Mathematics, 42(8), 358 (2023)

    work page 2023

  2. [1]

    Com- munications on Applied Mathematics and Computation, 5(4), 1564-1583 (2023)

    Ahmadi-Asl, S., An efficient randomized fixed-precision algorithm for tensor singular value decomposition. Com- munications on Applied Mathematics and Computation, 5(4), 1564-1583 (2023)

    work page 2023

  3. [2]

    Ahmadi-Asl, M.G Asante-Mensah, A

    S. Ahmadi-Asl, M.G Asante-Mensah, A. Cichocki, A-H. Phan, I. Oseledets, J. Wang, Fast cross tensor approxima- tion for image and video completion, Signal Processing, 213, 109121 (2023)

    work page 2023

  4. [3]

    H., Hachimi, A

    Bentbib, A. H., Hachimi, A. E., Jbilou, K., Ratnani, A. A., Tensor Regularized Nuclear Norm Method for Image and Video Completion, J. Optim. Theory Appl. 192(2) 401-425 (2022)

    work page 2022

  5. [4]

    P., Najafi-Kalyani, M., A preconditioning technique in conjunction with Krylov subspace methods for solving multilinear systems

    Beik, F. P., Najafi-Kalyani, M., A preconditioning technique in conjunction with Krylov subspace methods for solving multilinear systems. Appl. Math. Lett. 116 107051 (2021)

    work page 2021

  6. [5]

    T product Tensors Part I: Inequalities

    Chang, S. Y., Wei, Y., T product tensors part I: Inequalities. arXiv preprint arXiv:2107.06285 (2021)

    work page Pith review arXiv 2021

  7. [6]

    Journal of Mathematical Modeling, 10(1), 21-37 (2022)

    El Guide, M., El Ichi, A., Jbilou, K., Discrete cosine transform LSQR methods for multidimensional ill-posed problems. Journal of Mathematical Modeling, 10(1), 21-37 (2022). 21

    work page 2022

  8. [7]

    Tensor GMRES and Golub-Kahan Bidiagonalization methods via the Einstein product with applications to image and video processing

    Guide, M. E., Ichi, A. E., Beik, F. P., Jbilou, K., Tensor GMRES and Golub-Kahan Bidiagonalization methods via the Einstein product with applications to image and video processing. arXiv preprint arXiv:2005.07458 (2020)

    work page Pith review arXiv 2020

Show all 28 references
  1. [8]

    El Guide, M. O. H. A. M. E. D., El Ichi, A. L. A. A., Jbilou, K., Sadaka, R., On tensor GMRES and Golub-Kahan methods via the T-product for color image processing, Electron. J. Linear Algebra. 37 524-543 (2021)

    2021

  2. [9]

    Linear Algebra

    Elden, L., Ahmadi-Asl, S., Solving bilinear tensor least squares problems and application to Hammerstein identifi- cation, Numer. Linear Algebra. Appl. 26 (2) (2019) e2226

    2019

  3. [10]

    Fusiello, A., A matter of notation: Several uses of the Kronecker product in 3D computer vision, Pattern Recognit. Lett. 28(15), 2127-2132 (2007)

    2007

  4. [11]

    Jin, H., Bai, M., Ben´ ıtez, J., Liu, X., The generalized inverses of tensors and an application to linear models. Comput. Math. Appl. 74(3), 385-397 (2017)

    2017

  5. [12]

    E., Braman, K., Hao, N., Hoover, R

    Kilmer, M. E., Braman, K., Hao, N., Hoover, R. C., Third-order tensors as operators on matrices: A theoretical and computational framework with applications in imaging. SIAM J. Matrix Anal. Appl. 34(1), 148-172 (2013)

    2013

  6. [13]

    E., Martin, C

    Kilmer, M. E., Martin, C. D., Factorization strategies for third-order tensors, Linear Algebra Appl. 435(3), 641-658 (2011)

    2011

  7. [14]

    A., An efficient color/grayscale image encryption scheme based on hybrid chaotic maps

    Khalil, N., Sarhan, A., Alshewimy, M. A., An efficient color/grayscale image encryption scheme based on hybrid chaotic maps. Opt. Laser Technol. 143, 107326 (2021)

    2021

  8. [15]

    G., Bader, B

    Kolda, T. G., Bader, B. W., Tensor decompositions and applications. SIAM Rev. 51(3) 455-500 (2009)

    2009

  9. [16]

    Tensor-tensor product toolbox

    Lu, C. Tensor-tensor product toolbox. arXiv preprint arXiv:1806.07247 (2018)

    2018 arXiv

  10. [17]

    Pattern Anal

    Lu, C., Feng, J., Chen, Y., Liu, W., Lin, Z., Yan, S., Tensor robust principal component analysis with a new tensor nuclear norm, IEEE Trans. Pattern Anal. Mach. Intell. 42(4) 925-938 (2019)

    2019

  11. [18]

    D., Shafer, R., LaRue, B., An order-p tensor factorization with applications in imaging, SIAM J

    Martin, C. D., Shafer, R., LaRue, B., An order-p tensor factorization with applications in imaging, SIAM J. Sci. Comput. 35(1) A474-A490 (2013)

    2013

  12. [19]

    Asante-Mensah, S

    MG. Asante-Mensah, S. Ahmadi-Asl, A. Cichocki, Matrix and tensor completion using tensor ring decomposition with sparse representation, Machine Learning: Science and Technology 2(3), 035008 (2021)

    2021

  13. [21]

    Qi, L., Luo, Z., Tensor analysis: spectral theory and special tensors, Society for Industrial and Applied Mathematics (2017)

    2017

  14. [22]

    E., Miller, E

    Semerci, O., Hao, N., Kilmer, M. E., Miller, E. L., Tensor-based formulation and nuclear norm regularization for multienergy computed tomography, IEEE Trans. Image Process. 23(4) 1678-1693 (2014)

    2014

  15. [23]

    Neural Inf

    Settles, B., Craven, M., Ray, S., Multiple-instance active learning, Adv. Neural Inf. Process. Syst. 20 (2007)

    2007

  16. [24]

    Moore–Penrose inverse of tensors via Einstein product

    Sun, L., Zheng, B., Bu, C., Wei, Y. Moore–Penrose inverse of tensors via Einstein product. Linear Multilinear Algebra. 64(4) 686-698 (2016)

    2016

  17. [25]

    W., Xu, X., Iterative algorithms for solving some tensor equations

    Wang, Q. W., Xu, X., Iterative algorithms for solving some tensor equations. Linear Multilinear Algebra. 67(7), 1325-1349 (2019)

    2019

  18. [26]

    arXiv preprint arXiv:1312.1254 (2013)

    Xu, Y., Hao, R., Yin, W., Su, Z., Parallel matrix factorization for low-rank tensor completion. arXiv preprint arXiv:1312.1254 (2013)

    2013 arXiv

  19. [27]

    Signal Process

    Zhang, Z., Aeron, S., Exact tensor completion using t-SVD, IEEE Trans. Signal Process. 65(6) 1511-1526 (2016)

    2016

  20. [28]

    IEEE Comput

    Zhang, Z., Ely, G., Aeron, S., Hao, N., Kilmer, M., Novel methods for multilinear data completion and de-noising based on tensor-SVD, Proc. IEEE Comput. Soc. Conf. Comput. Vis. Pattern Recognit. 3842-3849 (2014)

    2014

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.