Pith. sign in

REVIEW 4 major objections 5 minor 107 references

SARATR-X-v2 establishes that a SAR pre-training target should be both approximately invariant to multiplicative speckle and multi-scale, and shows that fusing six fixed structural operators with learnable weights yields a target that reduce

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

A learnable-weighted fusion of six fixed, speckle-robust structural operators as the masked pre-training target transfers better than pixel targets on 10 of 12 SAR benchmarks.

T0 review reviewed 2026-08-01 challenge →

load-bearing objection Strong target-design idea and a clean ablation, but the headline results hinge on an ATRNet-STAR pre-training/evaluation overlap that the paper never explicitly rules out. the 4 major comments →

arxiv 2607.23238 v1 pith:J5MQO7YX submitted 2026-07-25 cs.CV

SARATR-X-v2: Scale-Aware Structural Pre-Training for SAR Foundation Models

Weijie Li , Yafei Song , Yongxiang Liu , Bowen Peng , Jie Zhou , Jingyuan Xia , Wei Yang , Tianpeng Liu
show 2 more authors
Zhen Liu Li Liu
This is my paper
classification cs.CV
keywords SAR pre-trainingmasked image modelingreconstruction target designspeckle stabilitymulti-scale structural targetlearnable cross-scale fusionself-supervised learningcoherent imaging
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

SARATR-X-v2 tries to settle what a pre-training target for synthetic aperture radar should be. The paper argues that a target must be approximately invariant to the multiplicative speckle inherent in coherent imaging, and must span the heterogeneous spatial scales that downstream tasks use. It constructs such a target by fusing six fixed structural operators—a blind-spot local aggregation and five directional log-ratio region contrasts—with learnable weights. The paper reports that this fused target is about two orders of magnitude less sensitive to synthetic speckle than raw pixel supervision, and that it achieves the best transfer results on ten of twelve SAR benchmarks spanning classification, detection, and segmentation. The deeper claim is that stability under speckle and downstream transfer are two facets of the same design principle.

Core claim

The paper's central discovery is that a SAR pre-training target built from fixed, physics-grounded extractors—defined so that multiplicative speckle cancels in log-ratio contrasts and is excluded by a blind spot—can be merged into a single supervision signal via learnable cross-scale weights, and that this signal transfers better than pixel-space reconstruction or any fixed single-scale structural target. The evidence includes a measured ~100x reduction in target drift under log-normal speckle perturbation, a strong rank correlation between target drift and downstream accuracy, and ablations showing the gain comes from the target design and its learned fusion rather than from the backbone or

What carries the argument

The load-bearing object is the fused structural target y = Σ α_s f_s(x̃). The finest branch f_1 is a blind-spot 3×3 mean that excludes the center pixel, preventing local speckle from entering the target; the five larger branches compute directional log-ratio contrasts between disjoint half-regions at radii 3, 5, 9, 13, and 17, so multiplicative speckle affecting both halves cancels in the difference. The alpha weights are learned softmax scalars, making the composition of scales adaptive. A standard masked L2 loss reconstructs this single target; the physics is carried by the target definition rather than by extra loss terms.

Load-bearing premise

The argument rests on SAR intensity being modeled as multiplicative unit-mean speckle and on the blind-spot and log-ratio operators canceling that speckle well enough; if real SAR noise is correlated, additive, or miscalibrated, the cancellation (and the transfer gain attributed to it) may weaken.

What would settle it

Pre-train the same backbone with the fused structural target and with pixel targets, but add correlated speckle or thermal noise to the real SAR pre-training images (not just synthetic log-normal speckle); if the transfer gap on held-out benchmarks collapses or reverses, the stability mechanism is not the driver. A more direct test: measure the speckle-induced drift of the trained encoder's own output features—not the fixed target maps—and check whether the ~100x reduction still holds.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

If this is right

  • Replacing pixel-space reconstruction with the fused structural target improves frozen-backbone transfer and fine-tuning across classification, detection, and segmentation benchmarks.
  • Pre-training targets with lower speckle drift tend to transfer better: across eight target settings, drift and 10-shot accuracy show a strong inverse rank correlation.
  • The learned fusion assigns roughly 90% of weight to the largest scale, yet the smaller branches add structured refinements at boundaries and strong scatterers; removing them hurts performance.
  • A target that is stable but too narrow in scale can transfer worse than pixel supervision, so stability alone is insufficient.
  • The gains are attributable to target design rather than backbone or data scale, since ablations hold the pre-training framework fixed.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The two-condition framework—speckle stability plus scale compatibility—likely transfers to other coherent imaging modalities such as ultrasound or OCT, where multiplicative noise and multi-scale anatomy matter; one could test the same target design there.
  • The drift measure could serve as a cheap, pre-training-free diagnostic: rank candidate targets by synthetic speckle drift to predict transfer, saving compute before full pre-training runs.
  • The scalar per-scale fusion could be extended to spatially varying weights, letting different image regions blend fine and coarse structure differently; this is a natural next test.
  • The paper only validates under synthetic log-normal speckle; an open question is whether the ~100x drift reduction and the transfer advantage survive correlated speckle, additive thermal noise, or radiometric miscalibration.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The paper proposes SARATR-X-v2, a self-supervised pre-training framework for SAR that replaces pixel-space masked reconstruction with a fused multi-scale structural target. The target is built from fixed SAR-motivated operators — a blind-spot 3×3 aggregation and five directional log-ratio region-contrast kernels spanning receptive fields from 3×3 to 35×35 — combined through learnable softmax weights. The authors argue that this target satisfies two conditions: physics-grounded stability under multiplicative speckle and semantic scale compatibility. They report state-of-the-art transfer results on 12 SAR benchmarks across classification, detection, and segmentation, and present a stability–transfer correlation (Spearman ρ = −0.93) based on eight pre-training targets. The paper includes ablations (Table IV) separating the contributions of the backbone, target definition, and fusion strategy, and an appendix with per-benchmark tables and implementation details.

Significance. If the results hold, the paper makes a valuable conceptual contribution: it articulates two measurable target-design criteria for SAR masked pre-training and instantiates them in a single, self-contained objective. The central ablation in Table IV is well designed: it isolates the target from the backbone by fixing iTPN and varying only the supervision, and it shows consistent gains from the structural target over pixel reconstruction. The manuscript is also unusually transparent in its limitations section (V-B), acknowledging the synthetic-only perturbation evidence. The detailed appendix tables and planned code release are strengths. However, two load-bearing issues substantially reduce current confidence: the possible pre-training/evaluation overlap for ATRNet-STAR (and other newly added datasets), and the mismatch between the abstract's 'learned representation' drift claim and the actual measurement of fixed target-operator drift. The stability-transfer correlation is also weaker than the text suggests, being based on eight in-sample runs with an admitted counterexample.

major comments (4)
  1. [Section IV-A / Appendix B / Tables II, VII, XIX] The pre-training corpus includes ATRNet-STAR with 108,569 images (Table II, Appendix B), while the downstream classification benchmark and the stability-transfer analysis use the ATRNet-STAR SOC-50 split (Table VII, Fig. 5, Table XIX). The text states only that the original SARATR-X collection is 'test-excluded'; it never states that the three newly added datasets, especially ATRNet-STAR, have their evaluation splits removed from pre-training. If the SOC-50 test images are among the 108,569 pre-training images, the +3.5 accuracy gain (Table VII) and the Fig. 5/Table XIX correlation are circular and inflated. Please provide a file-level overlap analysis between the pre-training images and each downstream test split, for ATRNet-STAR and also for SSDD and HRSID, which appear in both Table II and Table III. If overlap exists, the pre-training and/or evaluation must be redone with clean split
  2. [Abstract / Section IV-D / Appendix D, Eqs. (16)–(19)] The abstract and conclusion claim that the proposed target 'reduces perturbation drift in the learned representation by nearly two orders of magnitude.' The experiments in Fig. 4 and Appendix D measure the mean ℓ1 drift of the fixed target operators (and raw pixel patches), not the drift of the learned encoder's representations. No experiment perturbs the input and measures the change in encoder output embeddings. The claim as stated is therefore unsupported. Please either measure representation drift directly (e.g., ℓ1/ℓ2 change in pooled encoder features before vs. after speckle perturbation across checkpoints) or revise the wording in the Abstract and Section VI to say 'target feature drift' rather than 'learned representation drift.'
  3. [Fig. 5 / Table XIX / Section IV-D] The stability–transfer correlation is computed over exactly the eight pre-training runs that generated the drift measurements, with n=8, and the text acknowledges a concrete inversion (S2 has far lower drift than pixel supervision but lower accuracy). The claim that the stability ranking 'strongly predicts' or 'essentially predicts' transfer ranking is an overstatement for such a small, in-sample, single-seed-per-target comparison. The correlation is dominated by the pixel point and one multi-scale point. Please either add more targets, multiple seeds with error bars, and a hold-out/out-of-sample evaluation, or soften the conclusion to 'suggestive, with exceptions' and quantify the uncertainty (e.g., bootstrap confidence intervals for ρ).
  4. [Section III-B / Section V-B / Appendix D] The 'physics-grounded stability' argument is presented heuristically: the authors state that blind-spot aggregation and directional log-ratio contrast 'cancel' multiplicative speckle, but no first-order stability derivation is provided for the actual operators, which include sigmoid compression and log-of-sum followed by subtraction. For non-unit-mean, correlated, or non-log-normal speckle — or when the two half-regions have different mean backscatter — the cancellation is only approximate. The paper itself confines the empirical evidence to synthetic log-normal speckle (Section V-B). Since the transfer advantage is attributed to this stability property, please add a short analytical derivation of the claimed approximate invariance under the stated model, or add a real SAR speckle-perturbation experiment (e.g., multi-look or repeat-pass data), or explicitly restrict the stability claim t
minor comments (5)
  1. [Appendix D] The text refers to 'Sec. X' ('differ only in the reconstruction target (Sec. X)') — this is a placeholder and should be replaced with the actual section number.
  2. [Fig. 3] The caption in the provided version contains an extended run of corrupted byte sequences ('/uni00000018/...'), making the figure caption unreadable. Please ensure the final PDF renders the caption correctly.
  3. [Appendix A, Eq. (13)] Eq. (13) duplicates Eq. (9) with slightly different notation. Please unify the two presentations or remove the duplication.
  4. [Appendix D] The stability protocol says 'held-out SAR images' but does not specify which dataset those images come from. Please state the source and confirm that these images are disjoint from pre-training, especially given the ATRNet-STAR overlap concern above.
  5. [Table XIX] The Spearman p-value is reported as p<0.001 for n=8; please report the exact p-value and clarify whether any multiple-comparison correction was applied. Also specify whether the linear-probe hyperparameters were tuned per target or fixed across targets.

Circularity Check

1 steps flagged

ATRNet-STAR evaluation images may be present in pre-training; the paper's only test-exclusion guarantee covers the original fourteen datasets, not the newly added ATRNet-STAR, making the headline benchmark and the stability–transfer correlation partially circular if overlap exists.

specific steps
  1. fitted input called prediction [Section IV-A 'Pre-training Data'; Table II; Table VII; Appendix B 'Pre-training Data Collection'; Appendix C 'Image Classification']
    "Following SARATR-X, we construct an expanded pre-training data collection in Table II by augmenting its test-excluded fourteen-dataset collection with three recently released open-source SAR datasets, namely ATRNet-STAR [3], FAIR-CSAR [37], and M4-SAR [38]. Pre-training uses only SAR imagery and does not rely on manual annotations."

    The test-exclusion guarantee is attached only to the original fourteen datasets; ATRNet-STAR is newly added in full (108,569 images, Table II) and is simultaneously the headline classification benchmark (SOC-50, Table VII, and the transfer axis in Fig. 5/Table XIX). Nowhere does the paper state that the SOC-50 evaluation images were removed from the 108,569 pre-training images. If they were not removed, the reported +3.5 gain on ATRNet-STAR and the Fig. 5/Table XIX stability–transfer correlation are computed on images the encoder was pre-trained on: the 'prediction' reduces to re-scoring pre-training data rather than transfer, and the correlation's transfer axis is statistically forced by training on the evaluation set. This is a load-bearing gap in the evaluation protocol, not an external

full rationale

The core derivation is largely self-contained. The target y is defined by fixed structural extractors and learnable fusion weights (Eq. 9), the loss is a standard masked L2 reconstruction (Eq. 12), and the ablation chain in Table IV compares pixel vs. structural targets under fixed frameworks with external downstream tasks (MSTAR-SOC, SSDD, AIR-PolSAR-Seg-2.0); none of these steps is defined in terms of the claimed outcome. Self-citations of SAR-JEPA/SARATR-X provide background, but the paper re-validates the 'structural target beats pixel' premise in its own ablations, so the self-citation is not load-bearing. The abstract's 'drift in the learned representation' is actually measured on fixed target operators rather than trained encoders, and the stability–transfer correlation (Table XIX) is an in-sample Spearman over 8 runs with an admitted counterexample (S2); these are overclaims and weak evidence, but not circular reductions. The main circularity risk is the ATRNet-STAR overlap: the pre-training section only claims test-exclusion for the original fourteen datasets, while ATRNet-STAR is both pre-training data and the benchmark used for the headline classification gain and for the Fig. 5 stability–transfer analysis. If SOC-50 test images are among the 108,569 pre-training images, that portion of the central claim reduces to a fitted-input-called-prediction. Because 10 of 12 benchmarks and the ablation study remain external controls, the paper is only partially circular and conditionally so; score 5 reflects this partial, load-bearing gap.

Axiom & Free-Parameter Ledger

2 free parameters · 4 axioms · 0 invented entities

No new physical entities are introduced; the two conditions (physics-grounded stability, semantic scale compatibility) are conceptual constructs, not entities. The central claim rests on the multiplicative-speckle physics assumption, the approximate invariance of the chosen operators, and the untested assumption that target stability transfers to encoder stability. The learnable fusion weights and hand-chosen radii are the main fitted inputs.

free parameters (2)
  • Cross-scale fusion weights {w_s} = alpha = (0.030, 0.008, 0.005, 0.024, 0.023, 0.909) at epoch 1200 (Appendix A)
    Six scalar logits fitted jointly with the encoder; the paper's claim that 'the optimization discovers which scales matter' is read off these fitted values from a single representative run.
  • Receptive-field radii r_s = blind-spot 3x3 plus r_s in {3, 5, 9, 13, 17}
    Chosen by hand as a geometric progression; no sensitivity analysis over alternative radius sets is provided, so the scale coverage is a design input, not a derived quantity.
axioms (4)
  • domain assumption Multiplicative speckle model x_i = s_i * n_i with unit-mean noise
    The entire physics-grounded stability argument (Eq. 1 and Section III-B) assumes pixel intensity factorizes into backscatter and unit-mean multiplicative noise; real SAR also has correlated speckle, additive noise, and radiometric errors.
  • domain assumption Log-ratio and blind-spot operators are approximately speckle-invariant
    The paper asserts 'first-order stability' for log(mean) of half-regions and blind-spot sums without formalizing the residual: the log of a sum of log-normals has a variance-dependent bias that cancels in the difference only when the two half-regions have similar statistics.
  • domain assumption Reconstructing a stable target makes the encoder's representations stable and transferable
    The bridge from target stability to representation quality is the central premise of the method; its direct evidence is the 8-point in-sample correlation in Fig. 5/Table XIX, which includes one counterexample (S2).
  • domain assumption iTPN encoder plus L2 reconstruction yields transferable features without auxiliary losses
    Standard SSL assumption adopted from iTPN [36] and the SAR-JEPA line; validated only through downstream transfer results, not by a direct analysis of learned representations.

reviewed 2026-08-01 · how reviews work

0 comments
Cite this review

Pith. "Pith review of SARATR-X-v2: Scale-Aware Structural Pre-Training for SAR Foundation Models." pith.science (2026). https://pith.science/paper/J5MQO7YX

@misc{pith2026260723238,
  author       = {Pith},
  title        = {Pith review of: SARATR-X-v2: Scale-Aware Structural Pre-Training for SAR Foundation Models},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/J5MQO7YX}},
  note         = {Machine review of arXiv:2607.23238}
}
Share X Bluesky LinkedIn Reddit HN
read the original abstract

Masked image modeling has become a dominant paradigm for SAR pre-training, yet the design of the reconstruction target remains fundamentally unsettled. This article argues that a SAR pre-training target should satisfy two conditions to produce transferable representations: (i) physics-grounded stability, i.e., approximate invariance of the target operator to multiplicative speckle inherent in coherent imaging; and (ii) semantic scale compatibility, i.e., coverage of the heterogeneous spatial scales that downstream tasks demand. These two conditions are individually achievable but jointly difficult: physics-grounded stability favors fixed operators, while semantic scale compatibility favors data-driven composition. To this end, SARATR-X-v2 reconciles both within a single design. The target is constructed through fixed structural extractors spanning six receptive fields, from blind-spot local aggregation to directional log-ratio region contrast, and fused via learnable weights into one unified supervision signal for masked reconstruction. On twelve SAR benchmarks across classification, detection, and segmentation, SARATR-X-v2 achieves state-of-the-art transfer performance. Under synthetic speckle variation, the proposed target reduces perturbation drift in the learned representation by nearly two orders of magnitude relative to pixel-space supervision. Taken together, these results establish physics-grounded stability and semantic scale compatibility as a principled framework for pre-training target design under coherent imaging, and suggest that effective SAR pre-training is not about reconstructing more signal, but about reconstructing the right structural target.

Figures

Figures reproduced from arXiv: 2607.23238 by Bowen Peng, Jie Zhou, Jingyuan Xia, Li Liu, Tianpeng Liu, Weijie Li, Wei Yang, Yafei Song, Yongxiang Liu, Zhen Liu.

Figure 1
Figure 1. Figure 1: Perturbation-sensitive supervision and task￾scale mismatch limit SAR pre-training. Our framework addresses both through scale-aware structural pre￾training guided by physics-grounded stability and semantic scale compatibility. (a) SAR-specific perturbations destabilize pixel-space supervision, while downstream tasks require representations at different scales. (b) It constructs a fine￾to-coarse structural … view at source ↗
Figure 2
Figure 2. Figure 2: Overall framework for SARATR-X-v2 with scale-aware structural pre-training. The framework is structured as two stages. The bottom stage constructs the pre-training target: fixed multi-scale structural extractors spanning six receptive fields produce scale-specific responses, which are fused through learnable cross-scale weights into a single target y. Every operator is designed to provide robustness agains… view at source ↗
Figure 3
Figure 3. Figure 3: Comprehensive comparison of SARATR-X-v2 on twelve SAR benchmarks spanning classification, object detection, and semantic segmentation. The compared methods include representative SAR-specific models (e.g., SARATR-X [24], SARMAE [26], SARCLIP [63]), remote-sensing multi-modal framework (e.g., RingMoE [21], BabelRS [64], ViTP [65]), and task-specific methods tailored to individual benchmarks. Despite the div… view at source ↗
Figure 5
Figure 5. Figure 5: Stability–transfer relationship across different pre￾training targets. Each point corresponds to one supervision target used for pre-training, including pixel reconstruction, six fixed single-scale targets (S1–S6), and the proposed multi￾scale fusion target. The horizontal axis shows the mean ℓ1 difference under log-normal speckle perturbation at σ = 0.15, and the vertical axis reports 10-shot linear-probe… view at source ↗
Figure 4
Figure 4. Figure 4: Stability of different target features against synthetic speckle perturbation. Mean ℓ1 difference between target features extracted from pairs of speckle realizations of the same input image, shown as a function of perturbation strength. The multiplicative speckle noise follows a log-normal distribution with standard deviation σ. For each σ we compute the expectation over K=5 independent realizations and M… view at source ↗
Figure 6
Figure 6. Figure 6: Residual correction visualization of multi-scale fusion. For each sample (row), we show from left to right: the input SAR image x; the feature map from the last (largest) scale branch f6; the fused target feature y; and the absolute residual map |y −f6|. All feature maps are displayed with the viridis colormap except the residual map, which uses magma to enhance the visibility of small-magnitude correction… view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

107 extracted references · 7 linked inside Pith

  1. [1]

    Spaceborne synthetic aperture radar imaging algorithms: An overview,

    G.-C. Sun, Y . Liu, J. Xiang, W. Liu, M. Xing, and J. Chen, “Spaceborne synthetic aperture radar imaging algorithms: An overview,” IEEE Geosci. Remote Sens. Mag. , vol. 10, no. 1, pp. 161–184, 2021

    2021

  2. [2]

    Deep learning meets SAR: Concepts, models, pitfalls, and perspectives,

    X. Zhu, S. Montazeri, M. Ali, Y . Hua, Y . Wang, L. Mou, Y . Shi, F. Xu, and R. Bamler, “Deep learning meets SAR: Concepts, models, pitfalls, and perspectives,” IEEE Geosci. Remote Sens. Mag. , vol. 9, no. 4, pp. 143–172, 2021

    2021

  3. [3]

    ATRNet-STAR: A large dataset and benchmark toward remote sensing object recognition in the wild,

    Y . Liu, W. Li, L. Liu, J. Zhou, B. Peng, Y . Song, X. Xiong, W. Yang, T. Liu, Z. Liu, and X. Li, “ATRNet-STAR: A large dataset and benchmark toward remote sensing object recognition in the wild,”IEEE Transactions on Pattern Analysis and Machine Intelligence , vol. 48, no. 6, pp. 6735–6753, 2026

    2026

  4. [4]

    Advanced cognitive radar: Principles, systems, and essential applications,

    Y . Liu, W. Yang, X. Qiu, C. Yang, W. Jiang, F. Gini, and M. S. Greco, “Advanced cognitive radar: Principles, systems, and essential applications,” IEEE Aerospace and Electronic Systems Magazine , vol. 41, no. 4, pp. 68–81, 2026

    2026

  5. [5]

    Unpaired image deraining using reward-guided self-reinforcement strategy,

    Y . Chen, Y . Jin, X. Chen, Y . Wei, Z. Yan, and Y . Fu, “Unpaired image deraining using reward-guided self-reinforcement strategy,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), June 2026, pp. 1342–1354. IN PREPARATION FOR SUBMISSION 14

    2026

  6. [6]

    Unleashing channel potential: Space-frequency selection convolution for SAR object detection,

    K. Li, D. Wang, Z. Hu, W. Zhu, S. Li, and Q. Wang, “Unleashing channel potential: Space-frequency selection convolution for SAR object detection,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , 2024, pp. 17 323– 17 332

    2024

  7. [7]

    Dual information purification for lightweight SAR object detection,

    X. Yang, J. Sun, S. Duan, and D. Cheng, “Dual information purification for lightweight SAR object detection,” in Proceedings of the AAAI Conference on Artificial Intelligence , vol. 39, no. 9, 2025, pp. 9274– 9282

    2025

  8. [8]

    A tutorial on synthetic aperture radar,

    A. Moreira, P. Prats-Iraola, M. Younis, G. Krieger, I. Hajnsek, and K. P. Papathanassiou, “A tutorial on synthetic aperture radar,” IEEE Geoscience and Remote Sensing Magazine , vol. 1, no. 1, pp. 6–43, 2013

    2013

  9. [9]

    Foundation models in remote sensing: Evolving from unimodality to multimodality,

    D. Hong, C. Li, X. Li, G. Camps-Valls, and J. Chanussot, “Foundation models in remote sensing: Evolving from unimodality to multimodality,”IEEE Geoscience and Remote Sensing Magazine, 2026

    2026

  10. [10]

    Vision foundation models in remote sensing: A survey,

    S. Lu, J. Guo, J. R. Zimmer-Dauphinee, J. M. Nieusma, X. Wang, P. VanValkenburgh, S. A. Wernke, and Y . Huo, “Vision foundation models in remote sensing: A survey,” IEEE Geoscience and Remote Sensing Magazine, vol. 13, no. 3, pp. 190–215, 2025

    2025

  11. [11]

    Self-supervised learning in remote sensing: A review,

    Y . Wang, C. M. Albrecht, N. A. A. Braham, L. Mou, and X. X. Zhu, “Self-supervised learning in remote sensing: A review,” IEEE Geoscience and Remote Sensing Magazine , vol. 10, no. 4, pp. 213– 247, 2022

    2022

  12. [12]

    Robsense: A robust multi-modal foundation model for remote sensing with static, temporal, and incomplete data adaptability,

    M. K. Do, K. Han, P. Lai, K. T. Phan, and W. Xiang, “Robsense: A robust multi-modal foundation model for remote sensing with static, temporal, and incomplete data adaptability,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2025, pp. 7427–7436

    2025

  13. [13]

    Crossearth: Geospatial vision foundation model for domain generalizable remote sensing semantic segmentation,

    Z. Gong, Z. Wei, D. Wang, X. Hu, X. Ma, H. Chen, Y . Jia, Y . Deng, Z. Ji, X. Zhu, X. Yang, N. Yokoya, J. Zhang, B. Du, J. Yan, and L. Zhang, “Crossearth: Geospatial vision foundation model for domain generalizable remote sensing semantic segmentation,” IEEE Transactions on Pattern Analysis and Machine Intelligence , vol. 48, no. 5, pp. 5147–5164, 2026

    2026

  14. [14]

    A genealogy of foundation models in remote sensing,

    K. Lane and M. Karimzadeh, “A genealogy of foundation models in remote sensing,” arXiv preprint arXiv:2504.17177 , 2025

    arXiv 2025

  15. [15]

    A tutorial on speckle reduction in synthetic aperture radar images,

    F. Argenti, A. Lapini, T. Bianchi, and L. Alparone, “A tutorial on speckle reduction in synthetic aperture radar images,” IEEE Geoscience and Remote Sensing Magazine , vol. 1, no. 3, p. 6–35, 2013

    2013

  16. [16]

    DiffDet4SAR: Diffusion-based aircraft target detection network for SAR images,

    J. Zhou, C. Xiao, B. Peng, Z. Liu, L. Liu, Y . Liu, and X. Li, “DiffDet4SAR: Diffusion-based aircraft target detection network for SAR images,” IEEE Geosci. Remote Sens. Lett. , 2024

    2024

  17. [17]

    Discovering and explaining the noncausality of deep learning in SAR ATR,

    W. Li, W. Yang, L. Liu, W. Zhang, and Y . Liu, “Discovering and explaining the noncausality of deep learning in SAR ATR,” IEEE Geosci. Remote Sens. Lett. , vol. 20, pp. 1–5, 2023

    2023

  18. [18]

    Hierarchical disentanglement-alignment network for robust SAR vehicle recognition,

    W. Li, W. Yang, W. Zhang, T. Liu, Y . Liu, and L. Liu, “Hierarchical disentanglement-alignment network for robust SAR vehicle recognition,” IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens. , vol. 16, pp. 9661–9679, 2023

    2023

  19. [19]

    MaRS: A multi-modality very-high-resolution remote sensing foundation model with cross-granularity meta-modality learning,

    R. Yang, Y . Liu, H. Yan, Y . Zhou, Y . Fu, H. Luo, and Y . Zhong, “MaRS: A multi-modality very-high-resolution remote sensing foundation model with cross-granularity meta-modality learning,” in AAAI Conference on Artificial Intelligence , 2026

    2026

  20. [20]

    SkySense V2: A unified foundation model for multi-modal remote sensing,

    Y . Zhang, L. Ru, K. Wu, L. Yu, L. Liang, Y . Li, and J. Chen, “SkySense V2: A unified foundation model for multi-modal remote sensing,” in Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), October 2025, pp. 9136–9146

    2025

  21. [21]

    RingMoE: Mixture-of-modality-experts multi-modal foundation models for universal remote sensing image interpretation,

    H. Bi, Y . Feng, B. Tong, M. Wang, H. Yu, Y . Mao, H. Chang, W. Diao, P. Wang, Y . Yu, H. Peng, Y . Zhang, K. Fu, and X. Sun, “RingMoE: Mixture-of-modality-experts multi-modal foundation models for universal remote sensing image interpretation,” IEEE Transactions on Pattern Analysis and Machine Intelligence , vol. 48, no. 4, pp. 4388– 4405, 2026

    2026

  22. [22]

    FUSAR-KLIP: Towards multimodal foundation models for remote sensing,

    Y . Yang, X. Zhang, Q. Fang, J. Liu, Z. Ye, R. Li, L. Liu, and H. Wang, “FUSAR-KLIP: Towards multimodal foundation models for remote sensing,” arXiv preprint , 2025. [Online]. Available: https://arxiv.org/abs/2509.23927

    arXiv 2025

  23. [23]

    Predicting gradient is better: Exploring self-supervised learning for SAR ATR with a joint-embedding predictive architecture,

    W. Li, W. Yang, T. Liu, Y . Hou, Y . Li, Z. Liu, Y . Liu, and L. Liu, “Predicting gradient is better: Exploring self-supervised learning for SAR ATR with a joint-embedding predictive architecture,” ISPRS J. Photogramm. Remote Sens. , vol. 218, pp. 326–338, 2024

    2024

  24. [24]

    SARATR-X: Toward building a foundation model for SAR target recognition,

    W. Li, W. Yang, Y . Hou, L. Liu, Y . Liu, and X. Li, “SARATR-X: Toward building a foundation model for SAR target recognition,” IEEE Trans. Image Process., vol. 34, pp. 869–884, 2025

    2025

  25. [25]

    SUMMIT: A SAR foundation model with multiple auxiliary tasks enhanced intrinsic characteristics,

    Y . Du, Y . Chen, L. Huang, Y . Yang, P. Ghamisi, and Q. Du, “SUMMIT: A SAR foundation model with multiple auxiliary tasks enhanced intrinsic characteristics,” International Journal of Applied Earth Observation and Geoinformation , vol. 141, p. 104624, 2025

    2025

  26. [26]

    SARMAE: Masked autoencoder for sar representation learning,

    D. Liu, D. Wang, H. Wang, H. Chen, W. Jiang, Y . Cheng, H. Guo, W. Cui, and J. Zhang, “SARMAE: Masked autoencoder for sar representation learning,” arXiv preprint , 2025. [Online]. Available: https://arxiv.org/abs/2512.16635

    arXiv 2025

  27. [27]

    Croma: Remote sensing representations with contrastive radar-optical masked autoencoders,

    A. Fuller, K. Millard, and J. Green, “Croma: Remote sensing representations with contrastive radar-optical masked autoencoders,” Advances in Neural Information Processing Systems, vol. 36, pp. 5506– 5538, 2023

    2023

  28. [28]

    Anysat: One earth observation model for many resolutions, scales, and modalities,

    G. Astruc, N. Gonthier, C. Mallet, and L. Landrieu, “Anysat: One earth observation model for many resolutions, scales, and modalities,” in Proceedings of the Computer Vision and Pattern Recognition Conference, 2025, pp. 19 530–19 540

    2025

  29. [29]

    CrossEarth- SAR: A SAR-centric and billion-scale geospatial foundation model for domain generalizable semantic segmentation,

    Z. Ye, Z. Gong, N. Liao, X. Hu, D. Wang, H. Chen, C. Huang, Y . He, Y . Jia, X. Wang, H. Wang, X. Yang, and J. Yan, “CrossEarth- SAR: A SAR-centric and billion-scale geospatial foundation model for domain generalizable semantic segmentation,” arXiv preprint arXiv:2603.12008, 2026

    arXiv 2026

  30. [30]

    SAMBA: A scatter-guided masked bidirectional mamba foundation model for SAR target recognition,

    K. Wang, X. Pan, Z. Gu, X. Ai, Z. Xu, F. Zhao, and S. Xiao, “SAMBA: A scatter-guided masked bidirectional mamba foundation model for SAR target recognition,” arXiv preprint arXiv:2606.31668 , 2026

    Pith/arXiv arXiv 2026

  31. [31]

    A complex-valued SAR foundation model based on physically inspired representation learning,

    M. Wang, H. Bi, Y . Feng, L. Xin, S. Gong, T. Wang, Z. Yan, P. Wang, W. Diao, and X. Sun, “A complex-valued SAR foundation model based on physically inspired representation learning,” arXiv preprint,

  32. [32]

    A statistical and geometrical edge detector for SAR images,

    R. Touzi, A. Lopes, and P. Bousquet, “A statistical and geometrical edge detector for SAR images,” IEEE Trans. Geosci. Remote Sens. , vol. 26, no. 6, pp. 764–773, 1988

    1988

  33. [33]

    On detecting edges in speckle imagery,

    A. C. Bovik, “On detecting edges in speckle imagery,” IEEE Trans. Acoust. Speech Signal Process. , vol. 36, no. 10, pp. 1618–1627, 1988

    1988

  34. [34]

    SAR target recognition via supervised discriminative dictionary learning and sparse representation of the SAR-HOG feature,

    S. Song, B. Xu, and J. Yang, “SAR target recognition via supervised discriminative dictionary learning and sparse representation of the SAR-HOG feature,” Remote Sens., vol. 8, no. 8, p. 683, 2016

    2016

  35. [35]

    SAR- SIFT: a SIFT-like algorithm for SAR images,

    F. Dellinger, J. Delon, Y . Gousseau, J. Michel, and F. Tupin, “SAR- SIFT: a SIFT-like algorithm for SAR images,” IEEE Trans. Geosci. Remote Sens., vol. 53, no. 1, pp. 453–466, 2014

    2014

  36. [36]

    Integrally pre-trained transformer pyramid networks,

    Y . Tian, L. Xie, Z. Wang, L. Wei, X. Zhang, J. Jiao, Y . Wang, Q. Tian, and Q. Ye, “Integrally pre-trained transformer pyramid networks,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023, pp. 18 610–18 620

    2023

  37. [37]

    Fair-csar: A benchmark dataset for fine-grained object detection and recognition based on single-look complex sar images,

    Y . Wu, Y . Suo, Q. Meng, W. Dai, T. Miao, W. Zhao, Z. Yan, W. Diao, G. Xie, Q. Ke, Y . Zhao, K. Fu, and X. Sun, “Fair-csar: A benchmark dataset for fine-grained object detection and recognition based on single-look complex sar images,” IEEE Transactions on Geoscience and Remote Sensing , vol. 63, pp. 1–22, 2025

    2025

  38. [38]

    M4-sar: A multi-resolution, multi-polarization, multi-scene, multi-source dataset and benchmark for optical-sar fusion object detection,

    C. Wang, W. Lu, X. Li, J. Yang, and L. Luo, “M4-sar: A multi-resolution, multi-polarization, multi-scene, multi-source dataset and benchmark for optical-sar fusion object detection,” arXiv preprint,

  39. [39]

    CRTransSar: A visual transformer based on contextual joint representation learning for SAR ship detection,

    R. Xia, J. Chen, Z. Huang, H. Wan, B. Wu, L. Sun, B. Yao, H. Xiang, and M. Xing, “CRTransSar: A visual transformer based on contextual joint representation learning for SAR ship detection,” Remote Sens. , vol. 14, no. 6, p. 1488, 2022

    2022

  40. [40]

    Available: https://arxiv.org/abs/2505.10931

    [Online]. Available: https://arxiv.org/abs/2505.10931

    Pith/arXiv arXiv

  41. [41]

    Category- oriented localization distillation for SAR object detection and a unified benchmark,

    C. Wang, R. Ruan, Z. Zhao, C. Li, and J. Tang, “Category- oriented localization distillation for SAR object detection and a unified benchmark,” IEEE Trans. Geosci. Remote Sens. , vol. 61, pp. 1–14, 2023

    2023

  42. [42]

    Large-scale multi-class SAR image target detection dataset-1.0,

    J. Chen, Z. Huang, R. Xia, B. Wu, L. Sheng, L. Sun, and B. Yao, “Large-scale multi-class SAR image target detection dataset-1.0,” https: //radars.ac.cn/web/data/getData?dataType=MSAR, 2022

    2022

  43. [43]

    Synthetic SAR image generation using sensor, terrain and target models,

    A. Kusk, A. Abulaitijiang, and J. Dall, “Synthetic SAR image generation using sensor, terrain and target models,” in Proc. Eur. Conf. Synth. Aperture Radar, EUSAR 2016 . VDE, 2016, pp. 1–5

    2016

  44. [44]

    Improving SAR automatic target recognition models with transfer learning from simulated data,

    D. Malmgren-Hansen, A. Kusk, J. Dall, A. A. Nielsen, R. Engholm, and H. Skriver, “Improving SAR automatic target recognition models with transfer learning from simulated data,” IEEE Geosci. Remote Sens. Lett., vol. 14, no. 9, pp. 1484–1488, 2017

    2017

  45. [45]

    A SAR dataset for ATR development: the synthetic and measured paired labeled experiment (SAMPLE),

    B. Lewis, T. Scarnati, E. Sudkamp, J. Nehrbass, S. Rosencrantz, and E. Zelnio, “A SAR dataset for ATR development: the synthetic and measured paired labeled experiment (SAMPLE),” in Proc. SPIE Conf. Algorithms SAR Imagery , vol. 10987, 2019, pp. 39–54

    2019

  46. [46]

    The air force moving and stationary target recognition database,

    Air Force Research Laboratory, “The air force moving and stationary target recognition database,” https://www.sdms.afrl.af.mil/index.php? collection=mstar

  47. [47]

    SIVED: A SAR image dataset for vehicle detection based on rotatable bounding box,

    X. Lin, B. Zhang, F. Wu, C. Wang, Y . Yang, and H. Chen, “SIVED: A SAR image dataset for vehicle detection based on rotatable bounding box,” Remote Sens., vol. 15, no. 11, p. 2825, 2023

    2023

  48. [48]

    Complex SAR data,

    Sandia National Laboratories, “Complex SAR data,” https://www. sandia.gov/radar/complex-data/index.html. IN PREPARATION FOR SUBMISSION 15

  49. [49]

    OpenSARShip 2.0: A large-volume dataset for deeper interpretation of ship targets in Sentinel-1 imagery,

    B. Li, B. Liu, L. Huang, W. Guo, Z. Zhang, and W. Yu, “OpenSARShip 2.0: A large-volume dataset for deeper interpretation of ship targets in Sentinel-1 imagery,” inProc. SAR Big Data Era: Models Methods Appl. (BIGSARDATA), 2017, pp. 1–5

    2017

  50. [50]

    A SAR dataset of ship detection for deep learning under complex backgrounds,

    Y . Wang, C. Wang, H. Zhang, Y . Dong, and S. Wei, “A SAR dataset of ship detection for deep learning under complex backgrounds,” Remote Sens., vol. 11, no. 7, p. 765, 2019

    2019

  51. [51]

    SAR ship detection dataset (SSDD): Official release and comprehensive data analysis,

    T. Zhang, X. Zhang, J. Li, X. Xu, B. Wang, X. Zhan, Y . Xu, X. Ke, T. Zeng, H. Su et al. , “SAR ship detection dataset (SSDD): Official release and comprehensive data analysis,”Remote Sens., vol. 13, no. 18, p. 3690, 2021

    2021

  52. [52]

    HRSID: A high-resolution SAR images dataset for ship detection and instance segmentation,

    S. Wei, X. Zeng, Q. Qu, M. Wang, H. Su, and J. Shi, “HRSID: A high-resolution SAR images dataset for ship detection and instance segmentation,” IEEE Access, vol. 8, pp. 120 234–120 254, 2020

    2020

  53. [53]

    SAR- AIRcraft-1.0: High-resolution SAR aircraft detection and recognition dataset (in chinese),

    Z. Wang, Y . Kang, X. Zeng, Y . Wang, T. Zhang, and X. Sun, “SAR- AIRcraft-1.0: High-resolution SAR aircraft detection and recognition dataset (in chinese),” J. Radars, vol. 12, no. 4, pp. 906–922, 2023

    2023

  54. [54]

    AIR- SARShip-1.0: High-resolution SAR ship detection dataset,

    X. Sun, Z. Wang, Y . Sun, W. Diao, Y . Zhang, and K. Fu, “AIR- SARShip-1.0: High-resolution SAR ship detection dataset,” J. Radars, vol. 8, no. 6, pp. 852–862, 2019

    2019

  55. [55]

    Moving and stationary target acquisition and recognition (MSTAR) model-based automatic target recognition: Search technology for a robust ATR,

    J. R. Diemunsch and J. Wissinger, “Moving and stationary target acquisition and recognition (MSTAR) model-based automatic target recognition: Search technology for a robust ATR,” in Proc. 5th SPIE Conf. Algorithms SAR Imagery , vol. 3370, 1998, pp. 481–492

    1998

  56. [56]

    SEFEPNet: Scale expansion and feature enhancement pyramid network for SAR aircraft detection with small sample dataset,

    P. Zhang, H. Xu, T. Tian, P. Gao, L. Li, T. Zhao, N. Zhang, and J. Tian, “SEFEPNet: Scale expansion and feature enhancement pyramid network for SAR aircraft detection with small sample dataset,” IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens. , vol. 15, pp. 3365–3375, 2022

    2022

  57. [57]

    SARDet-100K: Towards open-source benchmark and toolkit for large- scale SAR object detection,

    Y . Li, X. Li, W. Li, Q. Hou, L. Liu, M.-M. Cheng, and J. Yang, “SARDet-100K: Towards open-source benchmark and toolkit for large- scale SAR object detection,” in Proc. Adv. Neural Inf. Process. Syst. (NeurIPS), 2024

    2024

  58. [58]

    FUSAR-Ship: Building a high-resolution SAR-AIS matchup dataset of Gaofen-3 for ship detection and recognition,

    X. Hou, W. Ao, Q. Song, J. Lai, H. Wang, and F. Xu, “FUSAR-Ship: Building a high-resolution SAR-AIS matchup dataset of Gaofen-3 for ship detection and recognition,” Sci. China Inf. Sci. , vol. 63, pp. 1–19, 2020

    2020

  59. [59]

    Air-polsar-seg-2.0: Polarimetric sar ground terrain classification dataset for large-scale complex scenes,

    W. Zhirui, Z. Liangjin, W. Yuelei, Z. Xuan, K. Jian, Y . Jian, and S. Xian, “Air-polsar-seg-2.0: Polarimetric sar ground terrain classification dataset for large-scale complex scenes,” Journal of Radars, vol. 14, no. 2, pp. 353–365, 2025

    2025

  60. [60]

    RSAR: Restricted state angle resolver and rotated sar benchmark,

    X. Zhang, X. Yang, Y . Li, J. Yang, M.-M. Cheng, and X. Li, “RSAR: Restricted state angle resolver and rotated sar benchmark,” in Proceedings of the Computer Vision and Pattern Recognition Conference, 2025, pp. 7416–7426

    2025

  61. [61]

    A dual-stream high resolution network: Deep fusion of gf-2 and gf- 3 data for land cover classification,

    B. Ren, S. Ma, B. Hou, D. Hong, J. Chanussot, J. Wang, and L. Jiao, “A dual-stream high resolution network: Deep fusion of gf-2 and gf- 3 data for land cover classification,” International Journal of Applied Earth Observation and Geoinformation , vol. 112, p. 102896, 2022

    2022

  62. [62]

    OpenEarthMap-SAR: A benchmark synthetic aperture radar dataset for global high-resolution land cover mapping [software and data sets],

    J. Xia, H. Chen, C. Broni-Bediako, Y . Wei, J. Song, and N. Yokoya, “OpenEarthMap-SAR: A benchmark synthetic aperture radar dataset for global high-resolution land cover mapping [software and data sets],” IEEE Geoscience and Remote Sensing Magazine , vol. 13, no. 4, pp. 476–487, 2025

    2025

  63. [63]

    SARVLM: A vision language foundation model for semantic understanding and target recognition in SAR imagery,

    Q. Ma, Z. Wang, W. Liu, X. Lu, B. Deng, P. Duan, X. Kang, and S. Li, “SARVLM: A vision language foundation model for semantic understanding and target recognition in SAR imagery,” arXiv preprint,

  64. [64]

    MCANet: A joint semantic segmentation framework of optical and SAR images for land use classification,

    X. Li, G. Zhang, H. Cui, S. Hou, S. Wang, X. Li, Y . Chen, Z. Li, and L. Zhang, “MCANet: A joint semantic segmentation framework of optical and SAR images for land use classification,” International Journal of Applied Earth Observation and Geoinformation , vol. 106, p. 102638, 2022

    2022

  65. [65]

    Visual instruction pretraining for domain-specific foundation models,

    Y . Li, Y . Zhang, W. Tang, Y . Dai, M.-M. Cheng, X. Li, and J. Yang, “Visual instruction pretraining for domain-specific foundation models,” arXiv preprint , 2025. [Online]. Available: https://arxiv.org/abs/2509.17562

    arXiv 2025

  66. [66]

    Available: https://arxiv.org/abs/2510.22665

    [Online]. Available: https://arxiv.org/abs/2510.22665

    Pith/arXiv arXiv

  67. [67]

    Unifying heterogeneous multi-modal remote sensing detection via language- pivoted pretraining,

    Y . Li, Y . Chen, Y . Li, M.-M. Cheng, X. Li, and J. Yang, “Unifying heterogeneous multi-modal remote sensing detection via language- pivoted pretraining,” in Int. Conf. Machin. Learn. (ICML) , 2026

    2026

  68. [68]

    A ConvNet for the 2020s,

    Z. Liu, H. Mao, C.-Y . Wu, C. Feichtenhofer, T. Darrell, and S. Xie, “A ConvNet for the 2020s,” in Proc. IEEE Comput. Soc. Conf. Comput. Vis. Pattern Recognit. (CVPR) , 2022, pp. 11 976–11 986

    2022

  69. [69]

    On the status of foundation models for SAR imagery,

    N. Inkawhich, “On the status of foundation models for SAR imagery,” arXiv preprint , 2025. [Online]. Available: https://arxiv.org/abs/2509. 21722

    2025

  70. [70]

    An image is worth 16x16 words: Transformers for image recognition at scale,

    A. Dosovitskiy, L. Beyer, A. Kolesnikov, D. Weissenborn, X. Zhai, T. Unterthiner, M. Dehghani, M. Minderer, G. Heigold, S. Gelly et al. , “An image is worth 16x16 words: Transformers for image recognition at scale,” arXiv preprint , 2020. [Online]. Available: https://arxiv.org/abs/2010.11929

    Pith/arXiv arXiv 2020

  71. [71]

    Semi-supervised SAR ATR via epoch- and uncertainty-aware pseudo-label exploitation,

    X. Zhang, Y . Luo, and L. Hu, “Semi-supervised SAR ATR via epoch- and uncertainty-aware pseudo-label exploitation,” IEEE Trans. Geosci. Remote Sens., vol. 61, pp. 1–15, 2023

    2023

  72. [72]

    Weakly contrastive learning via batch instance discrimination and feature clustering for small sample SAR ATR,

    Y . Zhai, W. Zhou, B. Sun, J. Li, Q. Ke, Z. Ying, J. Gan, C. Mai, R. D. Labati, V . Piuri, and F. Scotti, “Weakly contrastive learning via batch instance discrimination and feature clustering for small sample SAR ATR,” IEEE Trans. Geosci. Remote Sens. , vol. 60, pp. 1–17, 2022

    2022

  73. [73]

    Crucial feature capture and discrimination for limited training data SAR ATR,

    C. Wang, S. Luo, J. Pei, Y . Huang, Y . Zhang, and J. Yang, “Crucial feature capture and discrimination for limited training data SAR ATR,” ISPRS J. Photogramm. Remote Sens. , vol. 204, pp. 291–305, 2023

    2023

  74. [74]

    RS5M and GeoRSCLIP: A large-scale vision-language dataset and a large vision-language model for remote sensing,

    Z. Zhang, T. Zhao, Y . Guo, and J. Yin, “RS5M and GeoRSCLIP: A large-scale vision-language dataset and a large vision-language model for remote sensing,” IEEE Transactions on Geoscience and Remote Sensing, vol. 62, pp. 1–23, 2024

    2024

  75. [75]

    Optimal azimuth angle selection for limited SAR vehicle target recognition,

    L. Zhang, X. Leng, S. Feng, X. Ma, K. Ji, G. Kuang, and L. Liu, “Optimal azimuth angle selection for limited SAR vehicle target recognition,” Int. J. Appl. Earth Obs. Geoinf. , vol. 128, p. 103707, 2024

    2024

  76. [76]

    Reproducible scaling laws for contrastive language-image learning,

    M. Cherti, R. Beaumont, R. Wightman, M. Wortsman, G. Ilharco, C. Gordon, C. Schuhmann, L. Schmidt, and J. Jitsev, “Reproducible scaling laws for contrastive language-image learning,” in CVPR, 2023, pp. 2818–2829

    2023

  77. [77]

    Swin transformer: Hierarchical vision transformer using shifted windows,

    Z. Liu, Y . Lin, Y . Cao, H. Hu, Y . Wei, Z. Zhang, S. Lin, and B. Guo, “Swin transformer: Hierarchical vision transformer using shifted windows,” in Proc. IEEE Comput. Soc. Conf. Comput. Vis. Pattern Recognit. (CVPR), 2021, pp. 10 012–10 022

    2021

  78. [78]

    Remoteclip: A vision language foundation model for remote sensing,

    F. Liu, D. Chen, Z. Guan, X. Zhou, J. Zhu, Q. Ye, L. Fu, and J. Zhou, “Remoteclip: A vision language foundation model for remote sensing,” IEEE Transactions on Geoscience and Remote Sensing , vol. 62, pp. 1– 16, 2024

    2024

  79. [79]

    SAR target classification based on multiscale attention super-class network,

    D. Wang, Y . Song, J. Huang, D. An, and L. Chen, “SAR target classification based on multiscale attention super-class network,” IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens. , vol. 15, pp. 9004–9019, 2022

    2022

  80. [80]

    Gsdet: Gaussian splatting for oriented object detection,

    Z. Ding, J. Zhao, Y . Zhou, W.-l. Du, H. Zhu, and R. Yao, “Gsdet: Gaussian splatting for oriented object detection,” in Proceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence , 2025, pp. 900–908

    2025

Showing first 80 references.

This paper was first reviewed by deepseek-v4-flash on August 1, 2026.