Pith. sign in

REVIEW 3 major objections 6 minor 64 references

Theory and Fast Learned Solver for $\ell^1$-TV Regularization

T0 review · 3 major / 6 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read The paper proves that $\ell^1$-TV regularization recovers signals from Gaussian measurements once the sample count exceeds $(\sqrt{\Phi(s_r,s_g)}+t)^2+1$, with $\Phi$ depending jointly on sparsity and gradient sparsity, and unrolls a…

desk verdict The explicit sample-complexity bound for l1-TV is a real contribution; the PGM-ISTA convergence theory has a load-bearing gap, and one comparison remark overclaims against Tropp. read the letter →

arxiv 2412.03269 v1 pith:TN2HC6CQ submitted 2024-12-04 math.NA cs.NA

Xinling Liu , Jianjun Wang , Bangti Jin This is my paper · ORCID
classification math.NAcs.NA MSC 94A1290C2565K0568T07
keywords ℓ1-TVregularizationcompressedsensingsamplecomplexitytotalvariationproximalgradientmappingalgorithmunrollinglearnedsolverECGsignalrecovery
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's aim is to make the combined $\ell^1$-plus-total-variation penalty a theoretically grounded and practically fast way to recover signals that are both sparse and piecewise-constant. Its central theoretical result is a measurement bound: under Gaussian sensing, a signal with $s_r$ nonzero entries and $s_g$ nonzero adjacent differences is stably recovered once the number of measurements exceeds $(\sqrt{\Phi(s_r,s_g)}+t)^2+1$, with an explicit formula for $\Phi$. The same bound reduces to known TV and $\ell^1$ cases when one penalty is switched off, and in suitable parameter regimes it is smaller than either, meaning the joint model can need fewer measurements. On the algorithmic side, the paper constructs PGM-ISTA, proves global convergence, and unrolls it into LPGM-ISTA, which recovers ECG signals accurately in a handful of layers and far less time than standard iterative solvers.

What carries the argument

The load-bearing object is the descent cone $D(g,x)$ of the composite penalty $g(x)=\lambda_1\|x\|_1+\lambda_2\|Dx\|_1$ and its statistical dimension, the expected squared distance from a standard Gaussian vector to the polar cone. To bound the statistical dimension the authors choose vectors $z_1\in\partial\|x\|_1$ and $z_2\in\partial\|Dx\|_1$, form $\lambda_1z_1+\lambda_2D^{\top}z_2$, and use a sign-correlation expectation identity to evaluate $\mathbb{E}\|g-t(\lambda_1z_1+\lambda_2D^{\top}z_2)\|_2^2$; minimizing over $t$ yields $\Phi(s_r,s_g)$, and the inequality $w^2\le\delta$ converts the width bound into a measurement bound. The algorithmic machinery is the proximal gradient mapping update $x_{k+1}=T_{\lambda_2t}((1-t/u)x_k+(t/u)S_{\lambda_1u}(x_k-uA^{\top}(Ax_k-y)))$, where $S$ is soft thresholding and $T$ is the TV proximal map; unrolling this update with learnable weights $W_x$, $W_y$, $u$, $t$ yields the LPGM-ISTA network.

What would settle it

Run the exact setting of Theorem 3.2 for fixed $n$, $s_r$, $s_g$, $\lambda_1=\lambda_2=1$ and Gaussian $A$: compute $\Phi(s_r,s_g)$, take $m=\lfloor(\sqrt{\Phi(s_r,s_g)}+t)^2+2\rfloor$, and record the empirical success frequency over many draws; if it falls clearly below $1-e^{-t^2/2}$, the sample-complexity bound is false. Separately, compare $F(x_{k+1})-F(x^*)$ with the claimed bound $2r\epsilon+(t\kappa+\beta)u$ on a sequence of iterates with growing $\|x_k\|_2$; any finite $r$ violated by the curve would show the hidden bounded-domain assumption is doing real work.

Watch

Extended reading notes

Core claim

On its own terms, the central discovery is Theorem 3.2: if the unknown signal $x^*$ has regular sparsity $s_r=|S_R(x^*)|$ and gradient sparsity $s_g=|S_G(x^*)|$, and $A\in\mathbb{R}^{m\times n}$ has independent standard Gaussian rows with $y=Ax^*+e$ and $\|e\|_2\le\epsilon$, then the solution $\hat{x}$ of the constrained problem (1.4) obeys $\|x^*-\hat{x}\|_2\le 2\epsilon/(\sqrt{m-1}-\sqrt{\Phi(s_r,s_g)}-t)$ with probability at least $1-e^{-t^2/2}$, provided $m>(\sqrt{\Phi(s_r,s_g)}+t)^2+1$. Here $\Phi(s_r,s_g)$ is the closed-form quantity in (3.2) built from $n$, $\lambda_1$, $\lambda_2$, $s_r$ and $s_g$. The paper also claims that setting $\lambda_1=0$ recovers the TV-only sampling bound, that setting $\lambda_2=0$ gives an $\ell^1$ bound that can be sharper than the standard one for modest sparsity levels, and that PGM-ISTA converges globally for $u\in(0,2/\|A\|_2^2)$ and $t\in(0,u]$, with its unrolled version LPGM-ISTA outperforming iterative baselines on ECG data.

Load-bearing premise

The parameter-selection guarantee for PGM-ISTA silently assumes that the $\ell^1$ penalty term is only evaluated on vectors of bounded length, even though the actual objective allows vectors of any length, and the radius $r$ appearing in the error bound is never fixed by the problem data.

Editorial extensions

If this is right

  • If Theorem 3.2 is correct, the sample complexity of $\ell^1$-TV recovery is governed by $\Phi(s_r,s_g)$, so the measurement count decreases with stronger value sparsity or gradient sparsity and interpolates between the $\ell^1$ and TV regimes as $\lambda_1/\lambda_2$ varies.
  • The bound recovers the TV-only sampling number when $\lambda_1=0$ and gives an $\ell^1$ bound when $\lambda_2=0$; for $n=10^5$ and $s_r<5000$ the paper claims the $\ell^1$ case is sharper than the standard bound.
  • PGM-ISTA has a guaranteed convergent regime $u\in(0,2/\|A\|_2^2)$, $t\in(0,u]$, and the convergence analysis gives an explicit function-value error bound that shrinks as $u$ becomes smaller, at the cost of more iterations.
  • The unrolled LPGM-ISTA network achieves its best reported ECG recovery with very few layers (relative error $0.056$ at two layers and $0.039$ at ten layers) and computes in about $0.06$ to $0.10$ seconds, which the paper contrasts with iterative solvers needing hundreds or thousands of iterations.

Reading between the lines

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

  • Beyond the paper: the same descent-cone technique should transfer to other composite penalties of the form $\lambda_1\|x\|_1+\lambda_2\|\Psi x\|_1$ whenever the second operator's subdifferential has a tractable sign structure, giving sample-complexity formulas for overlapping-group and analysis sparsity models.
  • Beyond the paper: Theorem 4.2's hidden bounded-domain assumption means the stated parameter-selection criterion is not yet a theorem for the actual unbounded $\ell^1$ objective; a bounded-domain variant or an explicit bound on $\|x_k\|_2$ along the iterates would be needed to restore the guarantee.
  • Beyond the paper: since LPGM-ISTA is trained on a fixed distribution of ECG windows and Gaussian sensing matrices, its millisecond-level advantage is likely distribution-specific; the paper itself notes out-of-distribution failure, so the practical claim is best read as in-distribution acceleration rather than a replacement for convergent solvers in general.
  • Beyond the paper: the comparison table suggests the joint bound can be several times smaller than either single-penalty bound when $\lambda_1/\lambda_2=1$, implying that for signals like ECG the measurement budget itself could be reduced if the bound is tight; that implication is not tested in the paper.
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 / 6 minor

Summary. The manuscript studies the l1-TV regularized model (1.4)/(1.5) in compressed sensing. Its theoretical contribution is a sample-complexity bound for l1-TV recovery: Theorem 3.1 bounds the statistical dimension of the descent cone in terms of the sparsity level sr and gradient-sparsity level sg, and Theorem 3.2 converts this into an explicit measurement bound with a Gaussian-width probability estimate. The algorithmic contribution is PGM-ISTA, a proximal-gradient-mapping iteration for (1.5), for which the authors claim global convergence and a parameter-selection criterion, and whose unrolled version LPGM-ISTA is trained on ECG data. The paper reports numerical comparisons of l1, TV, and l1-TV and of LPGM-ISTA against iterative solvers.

Significance. The sample-complexity result is significant if correct: it gives a quantitative recovery guarantee for a combined l1+TV penalty that depends on both signal sparsity and gradient sparsity and reduces, in the limiting cases, to previously known TV and l1 bounds. The derivation is substantial and does not appear to be circular: no constant is fitted, and the Gaussian-width machinery is used in the standard way. The learned solver is also potentially useful, and the authors provide code and experiments on ECG signals. However, the algorithmic theory has load-bearing gaps: the proof of global convergence constructs only iteration-dependent step sizes, and the parameter-selection theorem uses a bounded-domain assumption that is false for the actual l1 term. These issues must be repaired before the algorithmic claims can be accepted.

major comments (3)
  1. [Section 6.3, proof of Theorem 4.1(a)] The proof of Theorem 4.1(a) does not establish the existence of a single pair (u,t) valid for the whole infinite sequence. Lemma 6.3, condition (6.11), gives a u0 depending on the current point x, with the denominator containing the nonzero coordinates xi. The induction in the proof of (6.18) only shows that for each finite k one can find a new u_k after x_{k+1} has been generated; nothing bounds these u_k away from zero. If a coordinate xi tends to 0 as k increases, the required 1/u can diverge, so the identification of Algorithm 1 with the proximal subgradient method on the entire trajectory is not justified. Consequently the claim that the sequence converges to a minimizer of (1.5) is unsupported, and the later statement S*∩F≠∅ (just after (4.5)) is also not proved. The proof also has a circular flavor: x_{k+1} depends on u, but u_{k+1} is chosen after x_{k+1} is observed.
  2. [Theorem 4.2 and Section 6.4] Theorem 4.2 as stated does not apply to problem (1.5). The theorem's statement does not include the assumption dom(g1)⊂{x:||x||2≤r}; that assumption is only mentioned in the introductory paragraph before the theorem, and for the actual choice g1(x)=λ1||x||1 the domain is all of Rn, so no finite r exists. The proof uses r in several essential places: inequality (4.6), Lemma 6.4, the estimate ||xk−x*u||2≤2r, and the final bound 2rε+(tκ+β)u. Since r is not defined by the problem data and the boundedness condition fails, the claimed parameter-selection guarantee is unsupported. This is a load-bearing issue because parameter selection is one of the paper's central algorithmic contributions.
  3. [Remark 1 and Remark 4] The numerical comparison in Remark 1 is incorrect. For n=10^5 and sr=5000, the right-hand side of (3.4) equals approximately 4.25×10^4, whereas the bound 2sr log(n/sr)+2sr from [52] equals approximately 4.00×10^4. Thus (3.4) is not a sharper upper bound in this regime, and the statement that (3.4) gives a tighter bound than [52] for sr<5000 is false. The same comparison is repeated in Remark 4 and in the contribution summary, so the claimed improvement over [52] for the l1 case is not supported. This does not invalidate Theorem 3.2 itself, but it requires correction of the stated comparison.
minor comments (6)
  1. [Introduction, after Eq. (1.4)] The reduction statement is swapped: if λ1=0 and λ2>0, (1.4) reduces to the TV problem (1.3), not the l1 problem (1.2); and if λ1>0 and λ2=0, it reduces to the l1 problem, not TV.
  2. [Theorem 3.2] The displayed error bound is missing parentheses: it should read ||x*−x^||2 ≤ 2ε/(√(m−1)−√Φ(sr,sg)−t). As printed, the denominator is ambiguous.
  3. [Proof of Theorem 3.1] In the sentence before (6.8), the text states |Sc_R| = sr−1; the correct value is n−sr. The subsequent algebra correctly uses n−sr, so this is a typo, but it should be fixed.
  4. [Eq. (6.20)] The first term after the inequality labeled (c) appears to contain an extra superscript 2 on the norm of (I_n−uA^TA)(x_k−x*); the subsequent bound uses the norm without squaring. Please clarify the displayed formula.
  5. [Section 4.2, learnable parameters] The parameter set ΘL is stated as {Wx, Wy, u, t}, but Wx = I_n−uA^TA is determined by u when A is fixed. The redundancy or intentional decoupling of Wx and u should be explained.
  6. [Algorithm 2] In Step 5, the dual update uses µ without a subscript: v^{k+1}=v^k+µ(x^{k+1}−z^{k+1}). This should presumably be µ_k or µ_{k+1} consistently with the other updates.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity; the main statistical-dimension and sample-complexity derivation is self-contained, with correctness gaps that are not circular.

full rationale

The central results, Theorems 3.1 and 3.2, are derived from known tools (statistical dimension, conic Gaussian width, Grothendieck's identity, and Tropp's recovery bound) without fitting any parameter to the predicted quantity. The bound Phi(sr,sg) is a closed-form expression in the sparsity and gradient-sparsity counts, and the recovery bound follows by substituting this estimate into an external Gaussian-width inequality. No input is renamed as a prediction, and no load-bearing claim is justified solely by a self-citation. The PGM-ISTA analysis contains a genuine non-circularity flaw: Theorem 4.1(a) proves only per-iterate existence of u_k via Lemma 6.3, without establishing a uniform positive lower bound, so the identification of Algorithm 1 with a globally convergent proximal subgradient method is not justified; likewise Theorem 4.2 silently assumes bounded dom(g1), which fails for g1 = lambda1 ||x||_1. These are correctness/rigor concerns, not instances of the derivation reducing to its own inputs. Remark 1's numerical comparison with [52] appears inaccurate, but that is also a correctness issue. The learned solver is trained by empirical risk minimization and the paper explicitly states its in-distribution limitation, so the experimental success is not presented as a theorem derived from the same data. Self-citations occur only in contextual or comparative remarks and are not load-bearing. Overall, no circular step that fits the enumerated patterns can be identified in the paper's own equations or citations.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

The sample-complexity derivation is parameter-free in the sense that no constant is fitted to data, but it is conditional on the regularization weights λ1, λ2. The algorithm section introduces an unstated bounded-domain assumption (r) that is not satisfied by the real problem, and the step sizes u,t are free algorithmic parameters. No new physical entities are introduced.

free parameters (4)
  • lambda1 (regularization weight) = 10^{-3} chosen by cross-validation in experiments; treated as model parameter in theory
    In the sample-complexity theorem, λ1 is a model parameter; in experiments, λ1 is selected by cross-validation from {1, 10^{-1}, ..., 10^{-5}} with λ2=1. The bound Φ(sr,sg) depends on the ratio λ1/λ2, so the theoretical guarantee is conditional on this choice.
  • lambda2 (TV weight) = 1 fixed in most experiments
    Set to 1 in Section 5.1 and 0.01 in Section 5.2; not derived from data.
  • u and t (PGM-ISTA step sizes) = ranges specified in Theorem 4.1, e.g., u in (0,2/||A||^2), t in (0,u]; chosen empirically in experiments
    Step sizes are free algorithmic parameters; the existence proof in Theorem 4.1(a) does not construct a concrete value, and experiments select them by cross-validation.
  • r (bounded-domain radius) = null
    Appears in Theorem 4.2's error bound but is not defined by the problem; the proof assumes dom(g1) is bounded by r, which is false for the l1 norm.
assumptions (5)
  • domain assumption Lemma 2.2 from [38]: prox of λ1||x||_1 + λ2||Dx||_1 equals S_{λ1}(T_{λ2}(x)).
    Used to make PGM-ISTA computationally tractable; the paper cites [38] but does not prove it.
  • standard math Grothendieck's identity (Lemma 6.1), standard Gaussian integral identity.
    Used in the proof of Theorem 3.1 to compute expectations over Gaussian vectors.
  • standard math Corollary 3.5 in [52]: robust recovery via Gaussian width of descent cone.
    Gives the template error bound used in Theorem 3.2.
  • standard math Theorem 6.1 in [18] linking statistical dimension to inf_t E[dist^2(g,t∂g(x))].
    Used in Section 6.1 to convert subdifferential distance to statistical dimension.
  • ad hoc to paper Bounded domain dom(g1) ⊂ {||x||_2 ≤ r} introduced in proof of Theorem 4.2.
    Not stated in Theorem 4.2 and not satisfied by the actual g1=λ1||·||_1; this is a silent assumption.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Theory and Fast Learned Solver for $\ell^1$-TV Regularization." pith.science (2026). https://pith.science/paper/TN2HC6CQ

@misc{pith2026241203269,
  author       = {Pith},
  title        = {Pith review of: Theory and Fast Learned Solver for $\ell^1$-TV Regularization},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/TN2HC6CQ}},
  note         = {Machine review of arXiv:2412.03269}
}
abstract

The $\ell^1$ and total variation (TV) penalties have been used successfully in many areas, and the combination of the $\ell^1$ and TV penalties can lead to further improved performance. In this work, we investigate the mathematical theory and numerical algorithms for the $\ell^1$-TV model in the context of signal recovery: we derive the sample complexity of the $\ell^1$-TV model for recovering signals with sparsity and gradient sparsity. Also we propose a novel algorithm (PGM-ISTA) for the regularized $\ell^1$-TV problem, and establish its global convergence and parameter selection criteria. Furthermore, we construct a fast learned solver (LPGM-ISTA) by unrolling PGM-ISTA. The results for the experiment on ECG signals show the superior performance of LPGM-ISTA in terms of recovery accuracy and computational efficiency.

Figures

Figures reproduced from arXiv: 2412.03269 by the authors.

Figure 1
Figure 1. The comparison of ℓ 1 , TV and ℓ 1 -TV under Gaussian sampling with a sampling ratio of 0.5. m = O( √ sn ln n) and the entries of A follow the standard Gaussian distribution. However, the bound is not optimal, and several works have improved the result [17, 18, 24, 25]. The TV problem (1.3) can be effectively solved using, e.g., proximal gradient descent algorithm and projected subgradient algorithm [5]. The use of … view at source ↗
Figure 2
Figure 2. For fixed λ2 = 1, plots of Φ(sr, sg) (sg ≤ 2sr) under multiple settings of n and λ1. Remark 3. Let n0 be an upper bound on the conic Gaussian width of the ℓ 1 or TV methods. We examine the case Φ(sr, sg) ≤ n0 more closely. Direct computation shows 2(n − sr) 2 λ1 λ2 2 + b λ1 λ2 + c ≥ 0, (3.5) with b = 4√ 2(n − sr)(n − 1 − sg) + (3n + 12 min{sr, sg})n0 and c = 4(n − 1 − sg) 2 + 4(2n + sg − 4)n0. (3.5) is a quadratic… view at source ↗
Figure 3
Figure 3. Neural network representation of PGM-ISTA. (a) PGM-ISTA; (b) LPGM-ISTA obtained by unfolding [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗
Figures from the paper (6 more)
Figure 4
Figure 4. Figure 4: A synthetic signal and its recovered one by the [PITH_FULL_IMAGE:figures/full_fig_p011_4.png]
Figure 5
Figure 5. Figure 5: Selections of λ1 for fixed λ2 = 1 under different situations. 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Sampling ratio 0 0.2 0.4 0.6 0.8 1 Success ratio ADMM Bound (3.6) Th 2.5 in [24] 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Sampling ratio 0 0.2 0.4 0.6 0.8 1 Success ratio A…
Figure 6
Figure 6. Figure 6: The theoretical bound (3.6) and the bound from Theorem 2.5 of [24] at different sparsity levels, with [PITH_FULL_IMAGE:figures/full_fig_p012_6.png]
Figure 7
Figure 7. Figure 7: The convergence of PGM-ISTA for fixed λ1 = λ2 = 0.01, (a) shows values of F(x 1000(u, t)) − F(x ∗ ); (b) shows the minimum iteration k such that F(x k (u, t))−F(x ∗ ) < 0.01; (c) shows values of F(x k (u, t))−F(x ∗ ) with multiple choices of u and t = 0.9u; (d) shows v…
Figure 8
Figure 8. Figure 8: The exact and recovered signals by three methods with a sampling ratio 0.5 and [PITH_FULL_IMAGE:figures/full_fig_p013_8.png]
Figure 9
Figure 9. Figure 9: The exact signal and the recovered ones by LADMM, S-FISTA, PGM-ISTA and LPGM-ISTA with [PITH_FULL_IMAGE:figures/full_fig_p014_9.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

64 extracted references · 61 canonical work pages

  1. [52]

    J. A. Tropp. Convex recovery of a structured signal from independent random linear measurements. In Sampling theory, a renaissance , Appl. Numer. Harmon. Anal., pages 67–101. Birkh¨ auser/Springer, Cham, 2015

    work page 2015

  2. [1]

    Amelunxen, M

    D. Amelunxen, M. Lotz, M. B. McCoy, and J. A. Tropp. Living on the edge: phase transitions in convex programs with random data. Inf. Inference, 3(3):224–294, 2014

    work page 2014

  3. [2]

    Angelosante, G

    D. Angelosante, G. B. Giannakis, and E. Grossi. Compressed sensing of time-varying signals. In 16th International Conference on Digital Signal Processing , pages 1–8, 2009

    work page 2009

  4. [3]

    S. W. Anzengruber and R. Ramlau. Morozov’s discrepancy principle for Tikhonov-type functionals with nonlinear operators. Inverse Problems, 26(2):025001, 17, 2010

    work page 2010

  5. [4]

    Barbano, v

    R. Barbano, v. Z. Kereta, A. Hauptmann, S. R. Arridge, and B. Jin. Unsupervised knowledge-transfer for learned image reconstruction. Inverse Problems, 38(10):104004, 28, 2022

    work page 2022

  6. [5]

    A. Beck. First-order methods in optimization . SIAM, Philadelphia, PA, 2017

    work page 2017

  7. [6]

    Beck and M

    A. Beck and M. Teboulle. A fast iterative shrinkage-thresholding algorithm for linear inverse problems. SIAM J. Imaging Sci. , 2(1):183–202, 2009

    2009

  8. [7]

    Beck and M

    A. Beck and M. Teboulle. Smoothing and first order methods: a unified framework. SIAM J. Optim. , 22(2):557–580, 2012

    work page 2012

Show all 64 references
  1. [8]

    S. Boyd, N. Parikh, E. Chu, B. Peleato, and J. Eckstein. Distributed optimization and statistical learning via the alternating direction method of multipliers. Foundations and Trends® in Machine learning, 3(1):1– 122, 2011

    2011

  2. [9]

    Cai and W

    J.-F. Cai and W. Xu. Guarantees of total variation minimization for signal recovery. Inf. Inference , 4(4):328–353, 2015

    2015

  3. [10]

    E. J. Cand` es, Y. C. Eldar, D. Needell, and P. Randall. Compressed sensing with coherent and redundant dictionaries. Appl. Comput. Harmon. Anal. , 31(1):59–73, 2011

    2011

  4. [11]

    E. J. Cand` es, J. Romberg, and T. Tao. Robust uncertainty principles: exact signal reconstruction from highly incomplete frequency information. IEEE Trans. Inform. Theory , 52(2):489–509, 2006

    2006

  5. [12]

    Chen and M

    G. Chen and M. Teboulle. Convergence analysis of a proximal-like minimization algorithm using Bregman functions. SIAM J. Optim. , 3(3):538–543, 1993

    1993

  6. [13]

    X. Chen, J. Liu, Z. Wang, and W. Yin. Theoretical linear convergence of unfolded ISTA and its practical weights and thresholds. In Advances in Neural Information Processing Systems , pages 9079–9089, 2018

    2018

  7. [14]

    Cherkaoui, J

    H. Cherkaoui, J. Sulam, and T. Moreau. Learning to solve TV regularized problems with unrolled algo- rithms. In Advances in Neural Information Processing Systems (NeurIPS) , pages 11513–11524, 2020

    2020

  8. [15]

    L. Condat. A direct algorithm for 1-d total variation denoising. IEEE Signal Proc. Lett., 20(11):1054–1057, 2013

    2013

  9. [16]

    L. Cui, L. Bai, Y. Wang, P. S. Yu, and E. R. Hancock. Fused lasso for feature selection using structural information. Pattern Recogn., 119:108058, 2021

    2021

  10. [17]

    S. Daei, F. Haddadi, and A. Amini. Sample complexity of total variation minimization. IEEE Signal Proc. Lett., 25(8):1151–1155, 2018

    2018

  11. [18]

    S. Daei, F. Haddadi, A. Amini, and M. Lotz. On the error in phase transition computations for compressed sensing. IEEE Trans. Inform. Theory , 65(10):6620–6632, 2019

    2019

  12. [19]

    Daubechies, M

    I. Daubechies, M. Defrise, and C. De Mol. An iterative thresholding algorithm for linear inverse problems with a sparsity constraint. Comm. Pure Appl. Math. , 57(11):1413–1457, 2004

    2004

  13. [20]

    D. L. Donoho. Compressed sensing. IEEE Trans. Inform. Theory , 52(4):1289–1306, 2006

    2006

  14. [21]

    D. L. Donoho. For most large underdetermined systems of linear equations the minimal l1-norm solution is also the sparsest solution. Comm. Pure Appl. Math. , 59(6):797–829, 2006

    2006

  15. [22]

    M. F. Duarte, M. A. Davenport, D. Takhar, J. N. Laska, T. Sun, K. F. Kelly, and R. G. Baraniuk. Single-pixel imaging via compressive sampling. IEEE Signal Proc. Mag. , 25(2):83–91, 2008. 21

    2008

  16. [23]

    Garudadri, P

    H. Garudadri, P. K. Baheti, S. Majumdar, C. Lauer, F. Mass´ e, J. van de Molengraft, and J. Penders. Artifacts mitigation in ambulatory ECG telemetry. In The 12th IEEE International Conference on e- Health Networking, Applications and Services , pages 338–344, 2010

    2010

  17. [24]

    Genzel, G

    M. Genzel, G. Kutyniok, and M. M¨ arz. ℓ1-analysis minimization and generalized (co-)sparsity: when does recovery succeed? Appl. Comput. Harmon. Anal. , 52:82–140, 2021

    2021

  18. [25]

    Genzel, M

    M. Genzel, M. M¨ arz, and R. Seidel. Compressed sensing with 1D total variation: breaking sample com- plexity barriers via non-uniform recovery. Inf. Inference, 11(1):203–250, 2022

    2022

  19. [26]

    Gregor and Y

    K. Gregor and Y. LeCun. Learning fast approximations of sparse coding. In Proceedings of the 27th International Conference on Machine Learning , pages 399–406, 2010

    2010

  20. [27]

    J. Hou, X. Liu, H. Wang, and K. Guo. Tensor recovery from binary measurements fused low-rankness and smoothness. Signal Process., 221:109480, 2024

    2024

  21. [28]

    J. Hou, J. Wang, F. Zhang, and J. Huang. One-bit compressed sensing via ℓp (0 < p <1)-minimization method. Inverse Problems, 36(5):055005, 35, 2020

    2020

  22. [29]

    Huang, Y

    J. Huang, Y. Jiao, B. Jin, J. Liu, X. Lu, and C. Yang. A unified primal dual active set algorithm for nonconvex sparse recovery. Stat. Sci. , 36(2):215–238, 2021

    2021

  23. [30]

    Y. Jiao, B. Jin, and X. Lu. A primal dual active set with continuation algorithm for the ℓ0-regularized optimization problem. Appl. Comput. Harmon. Anal. , 39(3):400–426, 2015

    2015

  24. [31]

    Jin and P

    B. Jin and P. Maass. Sparsity regularization for parameter identification problems. Inverse Problems , 28(12):123001, 70, 2012

    2012

  25. [32]

    Kabanava and H

    M. Kabanava and H. Rauhut. Analysis ℓ1-recovery with frames and Gaussian measurements. Acta Appl. Math., 140:173–195, 2015

    2015

  26. [33]

    Kabanava, H

    M. Kabanava, H. Rauhut, and H. Zhang. Robust analysis ℓ1-recovery from Gaussian measurements and total variation minimization. European J. Appl. Math. , 26(6):917–929, 2015

    2015

  27. [34]

    K. Li, H. Li, R. H. Chan, and Y. Wen. Selecting regularization parameters for nuclear norm-type mini- mization problems. SIAM J. Sci. Comput. , 44(4):A2204–A2225, 2022

    2022

  28. [35]

    X. Li, L. Mo, X. Yuan, and J. Zhang. Linearized alternating direction method of multipliers for sparse group and fused LASSO models. Comput. Statist. Data Anal. , 79:203–221, 2014

    2014

  29. [36]

    X. Li, D. Sun, and K.-C. Toh. On efficiently solving the subproblems of a level-set method for fused lasso problems. SIAM J. Optim. , 28(2):1842–1866, 2018

    2018

  30. [37]

    Liu and X

    J. Liu and X. Chen. ALISTA: Analytic weights are as good as learned weights in LISTA. In International Conference on Learning Representations (ICLR) , 2019

    2019

  31. [38]

    J. Liu, L. Yuan, and J. Ye. An efficient algorithm for a class of fused lasso problems. In Proceedings of the 16th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining , pages 323–332, 2010

    2010

  32. [39]

    X. Liu, J. Peng, J. Hou, Y. Wang, and W. Jianjun. Guaranteed matrix recovery using weighted nuclear norm plus weighted total variation minimization. Signal Process., 227:109706, 2025

    2025

  33. [40]

    Lustig, D

    M. Lustig, D. Donoho, and J. M. Pauly. Sparse MRI: The application of compressed sensing for rapid MR imaging. Magnetic Res. Med. , 58(6):1182–1195, 2007

    2007

  34. [41]

    Mo and Y

    Q. Mo and Y. Shen. A remark on the restricted isometry property in orthogonal matching pursuit. IEEE Trans. Inform. Theory, 58(6):3654–3656, 2012

    2012

  35. [42]

    Mohammadi

    M. Mohammadi. A compact neural network for fused lasso signal approximator. IEEE Trans. Cybern. , 51(8):4327–4336, 2019

    2019

  36. [43]

    Monga, Y

    V. Monga, Y. Li, and Y. C. Eldar. Algorithm unrolling: interpretable, efficient deep learning for signal and image processing. IEEE Signal Proc. Mag. , 38(2):18–44, 2021

    2021

  37. [44]

    G. B. Moody and M. R. G. The impact of the MIT-BIH arrhythmia database. IEEE Eng. Med. Biol. Mag., 20(3):45–50, 2001

    2001

  38. [45]

    J.-J. Moreau. Proximit´ e et dualit´ e dans un espace hilbertien.Bull. Soc. Math. France , 93:273–299, 1965. 22

    1965

  39. [46]

    Rapaport, E

    F. Rapaport, E. Barillot, and J. Vert. Classification of arrayCGH data using fused SVM. Bioinformatics, 24(13):i375–i382, 2008

    2008

  40. [47]

    R. T. Rockafellar. Convex analysis, volume No. 28 of Princeton Mathematical Series. Princeton University Press, Princeton, NJ, 1970

    1970

  41. [48]

    Scetbon, M

    M. Scetbon, M. Elad, and P. Milanfar. Deep K-SVD denoising. IEEE Trans. Imag. Proc. , 30:5944–5955, 2021

    2021

  42. [49]

    Sulam, A

    J. Sulam, A. Aberdam, A. Beck, and M. Elad. On multi-layer basis pursuit, efficient algorithms and convolutional neural networks. IEEE Trans. Pattern Anal. Mach. Intell. , 42(8):1968–1980, 2020

    1968

  43. [50]

    Tang and P

    L. Tang and P. X. K. Song. Fused lasso approach in regression coefficients clustering—learning parameter heterogeneity in data integration. J. Mach. Learn. Res. , 17:Paper No. 113, 23, 2016

    2016

  44. [51]

    Tibshirani, M

    R. Tibshirani, M. Saunders, S. Rosset, J. Zhu, and K. Knight. Sparsity and smoothness via the fused lasso. J. R. Stat. Soc. Ser. B Stat. Methodol. , 67(1):91–108, 2005

    2005

  45. [53]

    Vershynin

    R. Vershynin. High-dimensional probability, volume 47 of Cambridge Series in Statistical and Probabilistic Mathematics. Cambridge University Press, Cambridge, 2018. An introduction with applications in data science, With a foreword by Sara van de Geer

    2018

  46. [54]

    J. Wang, W. Fan, and J. Ye. Fused lasso screening rules via the monotonicity of subdifferentials. IEEE Trans. Pattern Anal. Mach. Intell. , 37(9):1806–1820, 2015

    2015

  47. [55]

    J. Wang, J. Huang, F. Zhang, and W. Wang. Group sparse recovery in impulsive noise via alternating direction method of multipliers. Appl. Comput. Harmon. Anal. , 49(3):831–862, 2020

    2020

  48. [56]

    J. Wang, F. Zhang, J. Huang, W. Wang, and C. Yuan. A nonconvex penalty function with integral convolution approximation for compressed sensing. Signal Process., 158:116–128, 2019

    2019

  49. [57]

    W. Wang, F. Zhang, and J. Wang. Low-rank matrix recovery via regularized nuclear norm minimization. Appl. Comput. Harmon. Anal. , 54:1–19, 2021

    2021

  50. [58]

    J. Wen, D. Li, and F. Zhu. Stable recovery of sparse signals via lp-minimization. Appl. Comput. Harmon. Anal., 38(1):161–176, 2015

    2015

  51. [59]

    J. Wen, Z. Zhou, J. Wang, X. Tang, and Q. Mo. A sharp condition for exact support recovery with orthogonal matching pursuit. IEEE Trans. Signal Process., 65(6):1370–1382, 2017

    2017

  52. [60]

    Wen and R

    Y.-W. Wen and R. H. Chan. Parameter selection for total-variation-based image restoration using discrep- ancy principle. IEEE Trans. Image Process., 21(4):1770–1781, 2012

    2012

  53. [61]

    B. Xin, Y. Kawahara, Y. Wang, and W. Gao. Efficient generalized fused lasso and its application to the diagnosis of alzheimer’s disease. In Proceedings of the 28-th AAAI Conference on Artificial Intelligence , volume 28, pages 2163–2169, 2014

    2014

  54. [62]

    Y. Yang, J. Sun, H. Li, and Z. Xu. ADMM-CSNet: a deep learning approach for image compressive sensing. IEEE Trans. Pattern Anal. Mach. Intell. , 42(3):521–538, 2020

    2020

  55. [63]

    Ye and X

    G.-B. Ye and X. Xie. Split Bregman method for large scale fused Lasso. Comput. Statist. Data Anal. , 55(4):1552–1569, 2011

    2011

  56. [64]

    Zhang, Y

    Y. Zhang, Y. Shen, R. Zhang, Y. Liu, Y. Guo, D. Deng, and I. D. Dinov. Numerical methods for computing the discrete and continuous Laplace transforms. Preprint, arXiv:2304.13204, 2023. 23

    2023 arXiv

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.