Pith. sign in

REVIEW 4 major objections 4 minor 28 references

Fast Physics-Driven Untrained Network for Highly Nonlinear Inverse Scattering Problems

T0 review · 4 major / 4 minor · reviewed 2026-08-02 · 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 a physics-driven untrained network solving in a severely truncated low-frequency Fourier basis reconstructs high-contrast microwave scattering targets in ~0.88 s, a ~100x speedup over prior untrained solvers, while st

desk verdict The speedup is real and measured; the 'high-fidelity' claim is not yet quantified, and the spectral-truncation assumption plus tuned CCO carries more weight than the paper acknowledges. read the letter →

arxiv 2602.13805 v1 pith:TCE3J5RX submitted 2026-02-14 cs.LG physics.comp-ph

Yutong Du , Zicheng Liu , Yi Huang , Bazargul Matkerim , Bo Qi , Yali Zong , Peixian Han This is my paper · ORCID
classification cs.LGphysics.comp-ph
keywords inversescatteringuntrainedneuralnetworksFourierbasisexpansioncontractionintegralequationcontrast-compensatedoperatormicrowaveimagingspectraldimensionalityreductionreal-timereconstruction
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 claims that the computational bottleneck of untrained neural-network inverse scattering solvers can be broken by moving the optimization from the spatial domain into a severely truncated low-frequency Fourier basis. The proposed PDF solver keeps only a few hundred Fourier coefficients of the induced current, uses the contraction integral equation to tame high-contrast nonlinearity, and adds a contrast-compensated operator plus a bridge-suppressing loss to correct the systematic blurring and edge roll-off that truncation causes. If true, this would make physics-driven, training-free microwave tomography fast enough for real-time medical, nondestructive-testing, and security-screening applications, while retaining the high fidelity and generalization-free nature of untrained solvers.

What carries the argument

The central object is the truncated Fourier-basis expansion of the induced current, J = F*_T(alpha), with only the four lowest-frequency MF×MF blocks of the 2D DFT retained (M0=4MF^2 coefficients, MF=7). This spectral parameterization turns a high-dimensional spatial inverse problem into a low-dimensional coefficient-estimation problem, where the forward operator is inherently low-pass because of the Green's function. Two supporting mechanisms carry the argument: the contraction integral equation, which replaces the contrast chi with the contractive modified contrast R = beta chi/(beta chi + 1) (beta=6) to weaken the nonlinear mapping; and the contrast-compensated operator, a self-guided pro

What would settle it

Take a target whose critical features are near the diffraction limit (e.g., an epsilon_r=8 annulus with a wall under one wavelength thick) and compare MF=7 with MF=12 reconstructions under identical hyperparameters; if MF=7 fails to resolve the wall yet MF=12 adds only artifacts instead of detail, the retained low-frequency subspace does not carry the information needed for sharp reconstruction, and the central claim collapses.

Watch

Extended reading notes

Core claim

The paper's central discovery is that a physics-constrained untrained network can converge to high-fidelity reconstructions in about 0.88 seconds—two orders of magnitude faster than existing untrained solvers (78–321 s)—by representing the induced current through a truncated discrete Fourier basis with MF=7 (only 196 low-frequency coefficients) and optimizing that compact coefficient vector with a fully connected network. The contraction integral equation reformulation makes the inverse mapping weakly nonlinear for high contrast, the contrast-compensated operator restores peak permittivity values lost at object boundaries due to spectral truncation, and the bridge-suppressing loss keeps clos

Load-bearing premise

The speedup rests on the claim that 196 low-frequency Fourier coefficients out of 4096 spatial pixels are enough to reconstruct high-contrast, sharp-edged objects and that all higher-frequency information is noise—demonstrated on the chosen benchmarks with parameters tuned on those same cases.

Editorial extensions

If this is right

  • Sub-second reconstruction (≈0.88 s) makes untrained physics-driven solvers viable for real-time microwave imaging, a ~100x reduction in runtime versus prior untrained solvers (78–321 s).
  • The solver maintains structural fidelity and target separation at high contrast (epsilon_r = 8) and under 1 dB SNR noise, where iterative baselines (SOM, FBE-CIE) and untrained baselines (uSOM, PDNN) degrade or fail.
  • Fourier truncation plus CIE contraction acts as an implicit regularizer: the optimizer does not diverge with increasing iterations and does not amplify high-frequency noise.
  • Adding the contrast-compensated operator and bridge-suppressing loss restores boundary permittivity values and removes false 'bridges' between adjacent scatterers, improving quantitative accuracy.
  • The method transfers to experimental scattering data, maintaining stable reconstructions across MF = 6–9, with MF = 7 giving the best balance.

Reading between the lines

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

  • If the low-pass spectral-compressibility premise holds generally, the same truncated-Fourier strategy could accelerate other untrained physics-driven inverse problems whose forward operators are band-limited, such as optical diffraction tomography or ultrasound inverse scattering.
  • A concrete testable extension: a coarse-to-fine schedule that grows MF during optimization might recover sharp edges even faster, using higher modes only where the physics residual demands them.
  • The robustness under antenna position uncertainties suggests the spectral parameterization may tolerate even larger calibration errors than the tested 3 mm; a direct experiment varying the perturbation distribution (e.g., correlated tilt errors rather than zero-mean jitter) would reveal whether the implicit filtering extends to systematic misalignment.
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

4 major / 4 minor

Summary. The paper proposes a physics-driven Fourier-spectral (PDF) solver for 2-D electromagnetic inverse scattering. The induced current is expanded in a truncated Fourier basis, reducing the optimization from the spatial grid to M0 = 4 M_F^2 low-frequency coefficients. The method combines a contraction integral equation (CIE), a fully connected network that updates Fourier coefficients, a contrast-compensated operator (CCO) to correct spectral edge roll-off, and a bridge-suppressing loss. Claims include sub-second reconstruction (~0.88 s), a ~100x speedup over untrained-network baselines (uSOM, PDNN), robustness to noise and antenna position uncertainty, and validation on simulated Austria-like profiles, additional geometries, and Fresnel experimental data. The runtime speedup is directly supported by Table 1, but reconstruction fidelity is presented almost entirely through selected images, with no quantitative error tables for the main comparisons, ablations, or experimental reconstructions.

Significance. If the fidelity claims are substantiated, the paper would be a useful step toward real-time microwave inverse scattering without supervised training data. The dimensionality-reduction idea is coherent: solving in a low-frequency Fourier subspace is a natural way to cut optimization cost, and the CIE-based physics loss is a reasonable response to high-contrast nonlinearity. The measured runtime advantage is credible and is a concrete, reproducible quantity. However, the central 'high-fidelity' half of the claim currently rests on qualitative figures and on hyperparameters selected on the same benchmark families used for evaluation. Because the paper explicitly acknowledges that spectral truncation causes systematic edge roll-off that must be patched by a hand-tuned CCO, the general spectral-compressibility premise needs stronger support. The paper would be significantly strengthened by quantitative error metrics, a proper tuning/validation split, and a clearer analysis of when the truncated Fourier representation can and cannot represent the target.

major comments (4)
  1. [Section 4.2.2, Fig. 7, Table 1] The central claim that PDF 'consistently achieves high-fidelity results' and 'outperforms state-of-the-art benchmarks' is not supported by any quantitative reconstruction error metric. Table 1 reports only runtime; Figures 7 and 10 are selected images. Please add numerical error tables (e.g., relative permittivity error, RMSE, SSIM) for all methods, all contrast levels, and all noise conditions, with statistics over multiple noise realizations. Without these, the fidelity claim is not quantitatively established.
  2. [Sections 4.1 and 4.2.1] The key hyperparameters (beta = 6, M_F = 7, lambda_1 = 1e-3, lambda_2 = 1e-5, lambda_3 = 1e-5, tau_B = 0.5, CCO gains tau = 3, eta_max = 0.1, delta = 0.5) are selected through sensitivity analysis on the Austria profile and then evaluated on Austria-like profiles and related experiments. This is effectively tuning on the test set. The claim that the method is 'consistently' high-fidelity across scenarios therefore needs independent validation: either a separate tuning set, a different evaluation geometry family, or an explicit demonstration that the chosen parameters perform well without per-benchmark adjustment.
  3. [Sections 2.3 and 2.4] The load-bearing premise that scattering measurements support only M0 = 4 M_F^2 low-frequency Fourier coefficients (M_F = 7) is asserted from the low-pass nature of the Green's function, not proven. The paper itself concedes that truncation produces systematic edge roll-off and requires a CCO with empirically tuned gains. No truncation-error analysis is given, and all evidence is restricted to the selected benchmark targets. Please provide a quantitative analysis of the truncation error (e.g., how much energy of the induced current lies in the retained modes for representative contrasts) and test the method on targets with genuinely high-frequency content, such as thin rods or fine periodic structures, to show that the premise holds beyond the chosen cases.
  4. [Section 5, Fig. 13] The experimental validation on Fresnel data is only qualitative. There is no quantitative comparison to the known cylinder permittivities, no error metrics, and no comparison with any baseline solver on the same experimental data. Since experimental data are central to the 'practical efficacy' claim, please add quantitative reconstruction errors (e.g., estimated permittivity vs. nominal values, regional error) and, if possible, compare with at least one iterative or untrained-network baseline on the FoamDielExt/FoamDielInt cases.
minor comments (4)
  1. [Fig. 12] The boxplot axes are unclear: the x-axis labels are missing, the y-axis text is garbled ('Relative Error (%)/s' and stray ASCII characters). Please redraw with clear axis labels and units, and include baseline comparisons or at least a reference error level.
  2. [Section 4.2.3] The sentence 'This performance bridge the gap' is ungrammatical; also 'without being constrained by generalization limits' overstates the case since robustness to distribution shift is not demonstrated beyond the tested configurations.
  3. [Fig. 2] The CCO workflow figure contains garbled inline text and unclear axes of the gain plots; please make the figure self-contained or refer the reader to the equations more explicitly.
  4. [Section 3.1] The phrase 'have been shown effective from the analysis in Section 4' is vague and does not identify which experiment supports the single-step gradient initialization. Please either provide a sentence explaining the empirical evidence or remove the claim.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation; spectral truncation and CCO are openly calibrated heuristics, not predictions derived from their own outputs.

full rationale

The paper's derivation chain is self-contained in the sense relevant to circularity. The forward scattering model (Eqs. 1-5), CIE reformulation (Eqs. 6-8), Fourier basis expansion (Eq. 9), and network losses (Eqs. 13-19) are defined from external physical and mathematical inputs; the optimized coefficient vector is fit to measured scattered fields through the physics-consistency losses. The speed claim in Table 1 is an externally measured runtime comparison (0.87-0.95 s vs 78-321 s), not an output derived from its own premise. The load-bearing spectral-compressibility assumption (Sec. 2.3, M_F=7) is a modeling premise, and the CCO and L_Bridge terms explicitly implement their intended effects (compensating truncation attenuation, suppressing low-gradient high-amplitude bridges), but the paper does not present these as first-principles predictions; they are calibrated components. Hyperparameters (β=6, M_F=7, λ's, τ=3, ηmax=0.1, δ=0.5) are openly selected via sensitivity analysis on the Austria profile and then used on Austria-like benchmarks, which is a generalization/calibration concern rather than a circular derivation: no equation is defined in terms of its own output and no fitted quantity is renamed as a prediction. Self-citations (e.g., refs. [18], [20]) are prior UNN baselines and are not invoked as load-bearing uniqueness theorems. Missing quantitative error metrics (Sec. 4.3 says 'quantitatively validated' but provides only visual comparisons) affect evidence strength, not circularity.

Assumptions & free parameters 7 free parameters · 5 assumptions · 0 invented entities

No new physical entities (particles, forces, dimensions) are introduced. The main burdens are the spectral-compressibility premise, the contractive CIE property imported from prior work, and a large set of hyperparameters tuned on the same evaluation benchmarks. These are algorithmic/domain assumptions rather than new physics.

free parameters (7)
  • Contraction parameter β = 6
    Selected via sensitivity analysis on the Austria profile (Section 4.1, Figs. 5-6); controls nonlinearity mitigation.
  • Fourier truncation order M_F = 7
    Selected via sensitivity analysis; controls number of retained modes M0 = 4 M_F^2 and reconstruction quality.
  • Loss weights λ1, λ2, λ3 = 1e-3, 1e-5, 1e-5
    Empirically set in Section 3.2; balance physics, boundary, TV, and bridge losses.
  • CCO gain parameters τ, η_max, δ = 3, 0.1, 0.5
    Hand-set in Section 2.4 to control contrast compensation strength and transition sharpness.
  • Bridge threshold τ_B = 0.5
    Empirically set in Section 3.2 for the bridge-suppressing loss.
  • Adam learning rate = 1e-2
    Fixed optimization setting reported in Section 3.3; affects convergence speed.
  • Iteration count k = up to 100
    Sensitivity analysis in Section 4.1 sweeps k = 10 to 100; final benchmarks use converged settings.
assumptions (5)
  • domain assumption Far-field scattering data are band-limited; the Green's function is effectively low-pass, so induced currents are compressible in a truncated Fourier basis.
    Invoked in Section 2.3 to justify dimensionality reduction; not proved, and M_F is chosen empirically.
  • domain assumption The CIE mapping χ → R = βχ/(βχ+1) is contractive for physically admissible media, taming high-contrast nonlinearity.
    Taken from CIE literature [21-23] and used in Section 2.2; not re-derived in this paper.
  • ad hoc to paper A single-step CSI gradient from zero gives a useful initial Fourier coefficient vector α^(0).
    Section 3.1 states this is 'shown effective from the analysis in Section 4'; no theoretical guarantee is provided.
  • ad hoc to paper The bridge-suppressing loss suppresses false connections between scatterers without suppressing genuine boundaries.
    Introduced as a heuristic in Section 3.2; threshold τ_B is chosen empirically and only image evidence supports it.
  • domain assumption Non-negative real contrast (Re{χ} ≥ 0, i.e., Re{ε_r} ≥ 1) is a valid physical bound for the media considered.
    Used to define L_Bound in Section 3.2; appropriate for passive dielectrics but not universal.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Fast Physics-Driven Untrained Network for Highly Nonlinear Inverse Scattering Problems." pith.science (2026). https://pith.science/paper/TCE3J5RX

@misc{pith2026260213805,
  author       = {Pith},
  title        = {Pith review of: Fast Physics-Driven Untrained Network for Highly Nonlinear Inverse Scattering Problems},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/TCE3J5RX}},
  note         = {Machine review of arXiv:2602.13805}
}
read the original abstract

Untrained neural networks (UNNs) offer high-fidelity electromagnetic inverse scattering reconstruction but are computationally limited by high-dimensional spatial-domain optimization. We propose a Real-Time Physics-Driven Fourier-Spectral (PDF) solver that achieves sub-second reconstruction through spectral-domain dimensionality reduction. By expanding induced currents using a truncated Fourier basis, the optimization is confined to a compact low-frequency parameter space supported by scattering measurements. The solver integrates a contraction integral equation (CIE) to mitigate high-contrast nonlinearity and a contrast-compensated operator (CCO) to correct spectral-induced attenuation. Furthermore, a bridge-suppressing loss is formulated to enhance boundary sharpness between adjacent scatterers. Numerical and experimental results demonstrate a 100-fold speedup over state-of-the-art UNNs with robust performance under noise and antenna uncertainties, enabling real-time microwave imaging applications.

Figures

Figures reproduced from arXiv: 2602.13805 by the authors.

Figure 1
Figure 1. Schematic of the 2-D inverse scattering configuration. [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Conceptual workflow of the contrast-compensated operator (CCO) demonstrating edge-restoration [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Computational pipeline of the proposed PDF solver, divided into initialization, neural network opti [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (11 more)
Figure 4
Figure 4. Figure 4: The network architecture of the proposed solver, where the input consists of two channels representing [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: A comprehensive parameter analysis by fixing [PITH_FULL_IMAGE:figures/full_fig_p009_5.png]
Figure 6
Figure 6. Figure 6: The reconstruction results of different hyperparameter [PITH_FULL_IMAGE:figures/full_fig_p010_6.png]
Figure 7
Figure 7. Figure 7: Reconstructed permittivity profiles for Austria targets under varying contrast levels ( [PITH_FULL_IMAGE:figures/full_fig_p012_7.png]
Figure 7
Figure 7. Figure 7: For low-contrast scenarios (ϵr = 2), when free of noise, all solvers recover the general geometry of the Austria profile. However, PDNN, SOM, and FBE-CIE frequently introduce non-physical “bridges” or false connections between the three components of the scatterer. As …
Figure 8
Figure 8. Figure 8: The reconstruction results of PDF and without the proposed contrast-compensated Operator (CCO) [PITH_FULL_IMAGE:figures/full_fig_p014_8.png]
Figure 9
Figure 9. Figure 9: The imaging results of PDF and without L Bound or L TV are compared under noise-free conditions and additive noise levels of SNR = 10dB, 5dB, and 1dB. electric interfaces. By applying the CCO as a post-processing step, we can adaptively compensate for this spectral-dom…
Figure 10
Figure 10. Figure 10: Comparison of imaging results for complex scatterers obtained by SOM, FBE-CIE, uSOM and PDF [PITH_FULL_IMAGE:figures/full_fig_p016_10.png]
Figure 11
Figure 11. Figure 11: The imaging results under antenna position uncertainties with no position perturbation, the best, [PITH_FULL_IMAGE:figures/full_fig_p017_11.png]
Figure 12
Figure 12. Figure 12: Boxplots of the relative errors under antenna position perturbations with Gaussian standard deviations [PITH_FULL_IMAGE:figures/full_fig_p018_12.png]
Figure 13
Figure 13. Figure 13: The imaging results of the proposed PDF solver using the experimental data ”FoamDielExt” and [PITH_FULL_IMAGE:figures/full_fig_p019_13.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

28 extracted references

  1. [1]

    Wiley, Hoboken, NJ, USA, 2018

    Xudong Chen.Computational Methods for Electromagnetic Inverse Scattering. Wiley, Hoboken, NJ, USA, 2018

    2018

  2. [2]

    Fused microwave and ultrasonic breast imaging within the framework of a joint variational bayesian approximation.IEEE Trans

    Yingying Qin, Thomas Rodet, and Dominique Lesselier. Fused microwave and ultrasonic breast imaging within the framework of a joint variational bayesian approximation.IEEE Trans. Antennas Propag., 70(12):12199–12211, 2022

    2022

  3. [3]

    Untrained neural network-based full-waveform inversion for breast sound speed imaging in ultrasound computed tomography.IEEE Trans

    Weicheng Yan, Qiude Zhang, Yun Wu, Zhaohui Liu, Hui Zhang, Zesong Wang, Jing Yuan, Mingyue Ding, Ming Yuchi, and Wu Qiu. Untrained neural network-based full-waveform inversion for breast sound speed imaging in ultrasound computed tomography.IEEE Trans. Instrum. Meas., 74:1–14, 2025

    2025

  4. [4]

    Microwave time reversal for nondestructive testing of buried small damage in composite materials.Inverse Probl., 40(4):045006, feb 2024

    Kang An, Changyou Li, Guoqian Long, and Jun Ding. Microwave time reversal for nondestructive testing of buried small damage in composite materials.Inverse Probl., 40(4):045006, feb 2024

    2024

  5. [5]

    Computational comparison and validation of point spread functions for optical microscopes.IEEE Trans

    Zicheng Liu, Yingying Qin, Jean-Claude Tinguely, and Krishna Agarwal. Computational comparison and validation of point spread functions for optical microscopes.IEEE Trans. Comput. Imaging, 2025

    2025

  6. [6]

    3d refractive index reconstruction from phaseless coherent optical microscopy data using multiple scattering-based inverse solvers—a study.Inverse Probl., 40(1):015003, nov 2023

    Yingying Qin, Ankit Butola, and Krishna Agarwal. 3d refractive index reconstruction from phaseless coherent optical microscopy data using multiple scattering-based inverse solvers—a study.Inverse Probl., 40(1):015003, nov 2023

    2023

  7. [7]

    A contrast source inversion method.Inverse Probl., 13(6):1607, 1997

    Peter M van den Berg and Ralph E Kleinman. A contrast source inversion method.Inverse Probl., 13(6):1607, 1997

    1997

  8. [8]

    Inversion of experimental multi-frequency data using the contrast source inversion method.Inverse Probl., 17(6):1611, 2001

    Richard F Bloemenkamp, Aria Abubakar, and Peter M van den Berg. Inversion of experimental multi-frequency data using the contrast source inversion method.Inverse Probl., 17(6):1611, 2001

    2001

Show all 28 references
  1. [9]

    Y. M. Wang and W. C. Chew. An iterative solution of the two-dimensional electromagnetic inverse scattering problem.Int. J. Imag. Syst. Technol., 1(1):100–108, 1989

    1989

  2. [10]

    Chew and Y.M

    W.C. Chew and Y.M. Wang. Reconstruction of two-dimensional permittivity distribution using the distorted Born iterative method.IEEE Trans. Med. Imaging, 9(2):218–225, 1990. 20

    1990

  3. [11]

    Subspace-based optimization method for solving inverse-scattering problems.IEEE Trans

    Xudong Chen. Subspace-based optimization method for solving inverse-scattering problems.IEEE Trans. Geosci. Remote Sens., 48(1):42–49, 2010

    2010

  4. [12]

    Deep-learning schemes for full-wave nonlinear inverse scattering prob- lems.IEEE Trans

    Zhun Wei and Xudong Chen. Deep-learning schemes for full-wave nonlinear inverse scattering prob- lems.IEEE Trans. Geosci. Remote Sens., 57(4):1849–1860, 2019

    2019

  5. [13]

    Teixeira, Che Liu, Arye Nehorai, and Tie Jun Cui

    Lianlin Li, Long Gang Wang, Fernando L. Teixeira, Che Liu, Arye Nehorai, and Tie Jun Cui. Deep- NIS: Deep neural network for nonlinear electromagnetic inverse scattering.IEEE Trans. Antennas Propag., 67(3):1819–1825, 2019

    2019

  6. [14]

    Physics-inspired convolutional neural network for solving full-wave inverse scattering problems.IEEE Trans

    Zhun Wei and Xudong Chen. Physics-inspired convolutional neural network for solving full-wave inverse scattering problems.IEEE Trans. Antennas Propag., 67(9):6138–6148, 2019

    2019

  7. [15]

    Prasad, and Krishna Agarwal

    Zicheng Liu, Mayank Roy, Dilip K. Prasad, and Krishna Agarwal. Physics-guided loss functions improve deep learning performance in inverse scattering.IEEE Trans. Comput. Imaging, 8:236–245, 2022

    2022

  8. [16]

    SOM-Net: Unrolling the subspace-based optimization for solving full-wave inverse scattering problems.IEEE Trans

    Yu Liu, Hao Zhao, Rencheng Song, Xudong Chen, Chang Li, and Xun Chen. SOM-Net: Unrolling the subspace-based optimization for solving full-wave inverse scattering problems.IEEE Trans. Geosci. Remote Sens., 60:1–15, 2022

    2022

  9. [17]

    Unrolled convolutional neural network for full-wave inverse scattering.IEEE Trans

    Yarui Zhang, Marc Lambert, Aur´ elia Fraysse, and Dominique Lesselier. Unrolled convolutional neural network for full-wave inverse scattering.IEEE Trans. Antennas Propag., 71(1):947–956, 2023

    2023

  10. [18]

    Quality-factor- inspired deep neural network solver for solving inverse scattering problems.IEEE Trans

    Yutong Du, Zicheng Liu, Miao Cao, Zupeng Liang, Yali Zong, and Changyou Li. Quality-factor- inspired deep neural network solver for solving inverse scattering problems.IEEE Trans. Geosci. Remote Sens., 63:1–13, 2025

    2025

  11. [19]

    Electromagnetic inverse scattering with an untrained som-net.IEEE Trans

    Rencheng Song, Meilan Li, Kuiwen Xu, Chang Li, and Xun Chen. Electromagnetic inverse scattering with an untrained som-net.IEEE Trans. Microwave Theory Tech., 70(11):4980–4990, 2022

    2022

  12. [20]

    Physics-driven neural network for solving electromagnetic inverse scattering problems.IEEE Trans

    Yutong Du, Zicheng Liu, Bazargul Matkerim, Changyou Li, Yali Zong, Bo Qi, and Jingwei Kou. Physics-driven neural network for solving electromagnetic inverse scattering problems.IEEE Trans. Antennas Propag., pages 1–1, 2025

    2025

  13. [21]

    A new integral equation method to solve highly nonlinear inverse scattering problems.IEEE Trans

    Yu Zhong, Marc Lambert, Dominique Lesselier, and Xudong Chen. A new integral equation method to solve highly nonlinear inverse scattering problems.IEEE Trans. Antennas Propag., 64(5):1788– 1799, 2016

    2016

  14. [22]

    Contraction integral equation for three-dimensional electromagnetic inverse scattering problems.J

    Yu Zhong and Kuiwen Xu. Contraction integral equation for three-dimensional electromagnetic inverse scattering problems.J. Imaging, 5(2), 2019

    2019

  15. [23]

    Fourier bases-expansion contraction integral equation for inver- sion highly nonlinear inverse scattering problem.IEEE Trans

    Kuiwen Xu, Lu Zhang, and Zhun Wei. Fourier bases-expansion contraction integral equation for inver- sion highly nonlinear inverse scattering problem.IEEE Trans. Microwave Theory Tech., 68(6):2206– 2214, 2020. 21

    2020

  16. [24]

    Guided image filtering

    He, Kaiming and Sun, Jian and Tang, Xiaoou. Guided image filtering. In Kostas Daniilidis, Petros Maragos, and Nikos Paragios, editors,Computer Vision – ECCV 2010, pages 1–14, Berlin, Heidel- berg, 2010. Springer Berlin Heidelberg

    2010

  17. [25]

    Guided image filtering.IEEE Trans

    Kaiming He, Jian Sun, and Xiaoou Tang. Guided image filtering.IEEE Trans. Pattern Anal. Mach. Intell., 35(6):1397–1409, 2013

    2013

  18. [26]

    An inverse algorithm for velocity reconstruction.Inverse Probl., 14(5):1345, 1998

    Tsili Wang and Michael L Oristaglio. An inverse algorithm for velocity reconstruction.Inverse Probl., 14(5):1345, 1998

    1998

  19. [27]

    Gibson.The Method of Moments in Electromagnetics

    W.C. Gibson.The Method of Moments in Electromagnetics. Chapman and Hall/CRC, New York, NY, USA, 2021

    2021

  20. [28]

    Free space experimental scattering database continuation: experimental set-up and measurement precision.Inverse Probl., 21(6):S117, 2005

    Jean-Michel Geffrin, Pierre Sabouroux, and Christelle Eyraud. Free space experimental scattering database continuation: experimental set-up and measurement precision.Inverse Probl., 21(6):S117, 2005. 22

    2005

Pith tools

Reviewed August 2, 2026 · model on record in the stance chip above.