Pith. sign in

REVIEW 5 major objections 5 minor 65 references

Exploring Generalizable Pre-training for Real-world Change Detection via Geometric Estimation

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

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

Pith's one-line read MatchCD performs geometric alignment and building change detection directly on large unregistered bi-temporal aerial images, reaching F1 scores up to 84.80 on distorted scenarios without manual registration.

desk verdict A plausible registration-plus-CD system whose headline results are under-evidenced: registration accuracy is never measured, and no registration-then-CD baseline is run on the same distorted inputs. read the letter →

arxiv 2504.14306 v1 pith:NKDXY6YY submitted 2025-04-19 cs.CV

Yitao Zhao , Sen Lei , Nanqing Liu , Heng-Chao Li , Turgay Celik , Qing Zhu This is my paper · ORCID
classification cs.CV
keywords changedetectionimageregistrationself-supervisedpre-trainingcontrastivelearninghomographyestimationremotesensingbuildingWarpCDdataset
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

Change detection in remote sensing normally assumes the two aerial images being compared are already aligned by a separate registration step. This paper claims that alignment does not need to be a manual or separate stage: a single framework, MatchCD, can estimate the geometric distortion between large unregistered bi-temporal images and then detect building changes in the same run, directly on full 6K-by-4K scenes. The authors support this with a self-supervised pre-training scheme that teaches an encoder to be robust to rotation and illumination changes, a training-free hierarchical matching procedure, and a new distorted benchmark called WarpCD on which MatchCD reports F1 scores between 71.85 and 84.80 without manual registration. If the claim holds, it removes a major bottleneck in real-world change detection workflows.

What carries the argument

The mechanism that carries the argument is a training-free hierarchical geometric estimation stage. Keypoint candidates are extracted by the GIM matcher from the original image pair and from two shallow feature maps (Conv1 and Layer1) of the pre-trained MatchCD encoder; the feature-map keypoints are re-localized to original-image coordinates by scale factors of 2 and 4, and the union of all candidates is used to solve a global 3-by-3 homography H via normalized DLT with RANSAC. Before change detection, the overlap boundary of the warped pair is computed and used to crop invalid regions. The downstream CD classifier then fuses MatchCD encoder features with FastSAM binary prior knowledge through cross-attention and decodes change tokens with a U-Net++ decoder.

What would settle it

Take a bi-temporal aerial pair with visible parallax (for example, tall buildings photographed from two different vantage points), run MatchCD's registration stage, and compare the estimated homography against the known ground-truth warp or measure the pixel error of matched keypoints after warping. If the global homography cannot explain the local displacements, the resulting change map will show systematic misregistration along building edges and rooftops, even where no change occurred.

Watch

Extended reading notes

Core claim

The central claim is that geometric alignment and change detection can be unified into one end-to-end pipeline, MatchCD, that works directly on large unregistered bi-temporal images. On the four WarpCD distortion scenarios, the framework simultaneously estimates the homography between the image pair and produces building change maps with F1 scores of 71.85, 84.80, 74.13, and 80.60, eliminating the manual registration step that conventional change detection requires. The authors further claim that the same pre-trained encoder, optimized by instance-level contrastive learning against class-agnostic instances generated with FastSAM, provides features that are reused for both registration and detection, and that the registration-detect pipeline outperforms existing CD baselines on the aligned WarpCD and WHU-CD benchmarks.

Load-bearing premise

The load-bearing premise is that a single global 3-by-3 homography can describe the distortion between the two large aerial images; the paper also states that it never directly measures registration accuracy, so this alignment assumption is not tested on its own.

Editorial extensions

If this is right

  • Change detection systems can accept unregistered bi-temporal aerial images directly, skipping the manual control-point selection and separate registration network that current workflows require.
  • Because the same encoder serves both registration and detection, the entire pipeline can be built without point-level registration labels, using only unlabeled imagery for the contrastive pre-training stage.
  • Processing full 6K-by-4K scenes in one pass instead of pre-segmented patches removes the content-mismatch failure mode that patch-based pipelines suffer under global distortion.
  • The overlap-boundary post-processing crops invalid regions after warping; ablation results show it raises F1 by roughly 2 to 3 points in three of the four distorted scenarios.

Reading between the lines

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

  • A natural stress test the paper leaves implicit: on scenes with parallax or non-planar terrain, replacing the single global homography with piecewise or depth-aware alignment would likely improve registration, and this could be evaluated directly once a registration-error benchmark exists.
  • Because the pre-training corpus is only about 51,000 class-agnostic instances from two scene-classification datasets, scaling instance-contrastive pre-training to larger unlabeled remote-sensing archives is a direct way to test whether the reported gains grow monotonically.
  • The same register-then-detect-on-full-scenes design should transfer to other bi-temporal tasks such as flood or deforestation mapping, where unregistered multi-date imagery is common; the main open question is whether the FastSAM prior remains useful for non-building objects.
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

5 major / 5 minor

Summary. The paper proposes MatchCD, a two-stage framework for change detection (CD) on large-scale, unregistered bi-temporal remote sensing images. In the first stage, FastSAM-generated class-agnostic instances from unlabeled aerial datasets (AID, UCMerced) are used in a DINO-style teacher-student contrastive pre-training to obtain a ResNet-50 encoder. In the second stage, the frozen encoder and the frozen GIM matcher are used for training-free hierarchical geometric estimation: keypoints are matched on original images and on Conv1/Layer1 feature maps, re-localized to original coordinates, fused, and used to estimate a global 3x3 homography by DLT+RANSAC; the post-change image is warped and an overlap boundary is computed. A change detection classifier then processes registered patches with the pre-trained encoder, FastSAM binary priors via cross-attention, and a U-Net++ decoder, followed by overlap-mask post-processing. The authors introduce WarpCD, a dataset with four scenarios and three distortion levels generated from WHU-CD regions, and report F1 scores of 71.85--84.80 on the four distorted scenarios (Table II), comparisons on aligned WarpCD scenarios (Tables III--VI), and WHU-CD results (Table VII).

Significance. If the central claim held, the framework would be practically valuable: it would remove the manual registration step from real-world CD workflows and allow direct processing of full 6Kx4K scenes. The release of WarpCD as a benchmark is a useful community contribution, and the training-free hierarchical matching idea is interesting. The paper is also refreshingly explicit about its limitation that registration accuracy is never directly evaluated (Section V-D). However, the current evidence does not support the central claim as strongly as the text suggests, because the main quantitative result (Table II) conflates registration quality with CD robustness and post-processing choices, and because the single-global-homography assumption for the Lv.3 viewpoint scenarios is left untested.

major comments (5)
  1. [Section V-D / Table II] The central claim -- that MatchCD simultaneously aligns and detects changes on unregistered 6Kx4K images -- is validated only indirectly through downstream CD metrics. Since WarpCD is generated from known synthetic warps, ground-truth alignment is available inside the pipeline; the paper should report direct registration accuracy, e.g., mean/probabilistic corner error, homography parameter error, or per-pixel alignment error on the known warps, for each scenario and level. Section V-D itself admits that 'such an approach cannot directly evaluate the performance of the registration task.' Without a direct measure, the Table II F1 values cannot distinguish successful geometric alignment from a CD model that is robust to residual misalignment or from the effect of overlap cropping.
  2. [Section IV-D2 / Table II] The unregistered-input evaluation (Table II) contains no registration-plus-CD baseline. The natural comparison is to run SIFT, LightGlue, or GIM to register the same distorted WarpCD inputs and then feed the registered images to a standard CD method such as SNUNet-CD or SEIFNet. Without such a two-stage baseline on the identical unregistered inputs, the 'simultaneous' joint claim is not tested against the most relevant alternative. The current Tables III--VI use already-aligned images and therefore do not measure the registration component of the framework.
  3. [Section III-C2 / Eq. (15)] The geometric estimation assumes that a single global 3x3 homography describes the distortion between the two large aerial images, including the Lv.3 viewpoint changes. For non-planar scenes or scenes with significant parallax, one homography cannot align the entire 6Kx3K image; residual local misalignment will be passed to the change detector. The paper should quantify the residual alignment error as a function of image location, e.g., by comparing the estimated homography against the known ground-truth warp on a grid, and discuss whether the scenes in WarpCD are sufficiently planar for this assumption. The current visual keypoint plots (Fig. 8) do not measure alignment quality.
  4. [Section V-C / Table X and Table III] Several reported results contradict the component claims and need explanation. In Table X, overlap detection improves Scenarios 1, 2, and 4 but decreases Scenario 3 F1 from 80.78 to 74.13 and IoU from 67.76 to 58.90; describing this as a 'slight decrease' is inaccurate. In Table III, MatchCD-base (75.87 F1) outperforms MatchCD (71.23 F1) on aligned Scenario 1, which is a negative-transfer result that is not discussed in the ablation or pre-training sections. Additionally, cropping by the overlap boundary changes the evaluation region, so F1/IoU before and after overlap detection are not directly comparable; the paper should report metrics on a fixed spatial region or also provide results on the common region for both settings.
  5. [Introduction, contribution list and Section II-A] The contribution claiming 'the first pioneering exploration of an end-to-end registration-change detection workflow' is not sufficiently differentiated from cited works [48] (R&D-Net, which integrates a registration net and a detection net) and [49] (a unified deep learning network for remote sensing image registration and change detection). The paper should clarify the precise distinction -- e.g., training-free vs. learned registration, full-scene matching vs. patch-level matching, or the specific use of pre-training -- or temper the novelty claim. Without this, the positioning of the contribution is unclear to the reader.
minor comments (5)
  1. [Section III-D, Eq. (17)] In Eq. (17), bin_T2 is computed as Θ_FastSAM(PT1_i), which appears to be a typo; it should likely be Θ_FastSAM(PT2_i).
  2. [Section IV-D3] The text refers to 'WarCD dataset' in the sentence introducing Fig. 10; this should be 'WarpCD' for consistency.
  3. [Section III-B, Eq. (6)] The definition of D(x,y) in Eq. (6) is notationally unclear: it writes a cross-entropy between softmax outputs but the expression -σ((x-C)/τ)·log(σ(y/τ)) does not show the expected sum over classes or the role of the clustering center C in a standard way. Please clarify the vector notation and the cross-entropy formulation.
  4. [Section IV-D2 / Tables III-VI] The tables report single-run results without standard deviations or repeated trials; given that the largest reported advantages over baselines are sometimes small (e.g., 0.16% F1 on WHU-CD), some indication of variance or statistical significance would help the reader judge the comparisons.
  5. [Section V-B1 / Fig. 13] The comparison of pre-training backbones is made using visual inspection of feature maps and keypoint matches; a quantitative measure of matching quality (e.g., number of correct matches under ground-truth warp) would make the choice of ResNet-50 over ViT-B more convincing.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the central pipeline is evaluated on external data; the admitted lack of direct registration metrics is an evaluation limitation, not a definitional reduction.

full rationale

The paper's derivation chain is self-contained against external data and benchmarks. The MatchCD encoder is pre-trained on unlabeled AID and UCMerced images using FastSAM-generated instances (Eqs. 1-7); the geometric stage is a training-free combination of a frozen GIM matcher with a RANSAC/DLT homography fit (Eqs. 8-15); and the change detector is trained on aligned WHU-CD labels with independent FastSAM masks. The WarpCD F1 values in Table II are genuine outputs of this full pipeline, not quantities fitted and then re-reported as predictions. The self-citations [32], [55], and [56] appear only as contextual descriptions of prior work and are not load-bearing; no uniqueness theorem or rejected alternative is smuggled in via a self-citation. The paper's own Section V-D admits that 'visual assessment and the performance of downstream change detection are employed to indirectly reflect the accuracy of the registration. However, such an approach cannot directly evaluate the performance of the registration task.' This is an evaluation gap: the known warps used to construct WarpCD would permit a direct pixel-level registration error, and the paper does not provide one. But that is a correctness/evidence limitation, not circularity, because no equation defines registration accuracy as the downstream F1 score, and Table II is not a renamed fit of the input warp. The absence of a pre-register-then-detect baseline is likewise an evidence concern rather than a reduction of the result to its inputs. No circular step can therefore be exhibited under the required standard.

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

The central claim rests on a pipeline of existing frozen models (FastSAM, GIM) plus a pre-trained ResNet-50 and a trained U-Net++ decoder. The main unstated premises are the global homography assumption, zero-shot transfer of natural-image models to aerial images, and the indirect evaluation of registration via CD metrics. No new physical entities are introduced.

free parameters (5)
  • Loss temperature tau (DINO-style) = not reported
    Temperature in Eq. (6); value is not given in the paper and must be tuned to reproduce the contrastive pre-training.
  • BCE positive-sample weight omega = not reported
    Weight for positive samples in Eq. (25), not reported; affects the F1/IoU trade-off of the CD classifier.
  • Mask filtering range for pre-training instances = 10% to 50% pixel portion
    Instances with pixel portion outside this range are discarded; hand-chosen threshold affects pre-training data distribution.
  • Number of attention heads = 8
    Selected by ablation (Fig. 12); other choices degrade F1/IoU, so the reported CD gains depend on this choice.
  • Patch partition size for downstream CD = not reported
    Patch size is never stated; it controls the token grid and the final change map stitching.
assumptions (5)
  • domain assumption Global 3x3 homography models the bi-temporal geometric distortion (Eq. 15)
    The geometric estimation solves a single H for the whole 6K-class scene; parallax or non-planar scene structure would violate this.
  • domain assumption GIM matcher and FastSAM transfer zero-shot to remote sensing imagery
    Pre-trained on natural images and internet videos; the paper assumes their keypoint matches and masks remain reliable on aerial imagery without fine-tuning.
  • ad hoc to paper Re-localization factors 2x and 4x exactly invert the encoder's downsampling
    Equations (11)-(12) multiply Conv1 and Layer1 keypoint coordinates by fixed factors, ignoring padding and boundary effects; if off, the merged keypoint set is biased.
  • domain assumption Downstream CD performance is a valid proxy for registration quality
    The paper's limitation section (Section V-D) states registration is not directly evaluated; all registration claims rely on CD metrics or visual inspection.
  • domain assumption WHU-CD labels used for training are independent of the WarpCD test regions
    The split between training 'remaining portion' and evaluation 'remaining regions' is not specified clearly, leaving potential train/test overlap.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Exploring Generalizable Pre-training for Real-world Change Detection via Geometric Estimation." pith.science (2026). https://pith.science/paper/NKDXY6YY

@misc{pith2026250414306,
  author       = {Pith},
  title        = {Pith review of: Exploring Generalizable Pre-training for Real-world Change Detection via Geometric Estimation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/NKDXY6YY}},
  note         = {Machine review of arXiv:2504.14306}
}
read the original abstract

As an essential procedure in earth observation system, change detection (CD) aims to reveal the spatial-temporal evolution of the observation regions. A key prerequisite for existing change detection algorithms is aligned geo-references between multi-temporal images by fine-grained registration. However, in the majority of real-world scenarios, a prior manual registration is required between the original images, which significantly increases the complexity of the CD workflow. In this paper, we proposed a self-supervision motivated CD framework with geometric estimation, called "MatchCD". Specifically, the proposed MatchCD framework utilizes the zero-shot capability to optimize the encoder with self-supervised contrastive representation, which is reused in the downstream image registration and change detection to simultaneously handle the bi-temporal unalignment and object change issues. Moreover, unlike the conventional change detection requiring segmenting the full-frame image into small patches, our MatchCD framework can directly process the original large-scale image (e.g., 6K*4K resolutions) with promising performance. The performance in multiple complex scenarios with significant geometric distortion demonstrates the effectiveness of our proposed framework.

Figures

Figures reproduced from arXiv: 2504.14306 by the authors.

Figure 1
Figure 1. Conventional CD methods vs. Our MatchCD. Illustration of the difference between conventional CD methods and our MatchCD in unregistered scenario. extensive data support for earth observation system. Based on this, the performance of various RS techniques is also comple￾mentarily enhanced from the multi-source data [5, 6, 7, 8, 9]. Conventional CD methods focus on revealing the change patterns between the multi-tempo… view at source ↗
Figure 3
Figure 3. Illustration of the hierarchical geometric estimation for bi-temporal [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figure 4
Figure 4. Illustration of downstream change detection workflow. [PITH_FULL_IMAGE:figures/full_fig_p006_4.png] view at source ↗
Figures from the paper (10 more)
Figure 5
Figure 5. Figure 5: Illustration of the dataset utilized for MatchCD pre-training. The samples are selected from the [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 6
Figure 6. Figure 6: Illustration of the three-level geometric distortions in the proposed [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]
Figure 7
Figure 7. Figure 7: Comparative performance of image matching algorithms on [PITH_FULL_IMAGE:figures/full_fig_p008_7.png]
Figure 8
Figure 8. Figure 8: Matching performance of the proposed MatchCD method in all distorted test scenarios in [PITH_FULL_IMAGE:figures/full_fig_p009_8.png]
Figure 9
Figure 9. Figure 9: Illustration of detailed change detection results of the proposed MatchCD method in all the distorted test scenarios on the registered [PITH_FULL_IMAGE:figures/full_fig_p011_9.png]
Figure 10
Figure 10. Figure 10: Illustration of full scenario change detection results of the proposed MatchCD method in all distorted test scenarios in [PITH_FULL_IMAGE:figures/full_fig_p011_10.png]
Figure 11
Figure 11. Figure 11: Comparison of different attention mechanisms applied in MatchCD [PITH_FULL_IMAGE:figures/full_fig_p012_11.png]
Figure 12
Figure 12. Figure 12: Ablation study of attention head number in MatchCD for change [PITH_FULL_IMAGE:figures/full_fig_p012_12.png]
Figure 14
Figure 14. Figure 14: Comparison of different self-supervised pre-training paradigms. [PITH_FULL_IMAGE:figures/full_fig_p013_14.png]
Figure 13
Figure 13. Figure 13: Visualization of latent features from different backbones. [PITH_FULL_IMAGE:figures/full_fig_p013_13.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

65 extracted references · 58 canonical work pages

  1. [48]

    R&D-Net: Integration of registration-net and detection-net for identifying building changes in high spatial- resolution remote sensing images,

    Z. Chen, T. Zheng, J. Leng, J. Zhang, H. Deng, X. Mi, and J. Yang, “R&D-Net: Integration of registration-net and detection-net for identifying building changes in high spatial- resolution remote sensing images,” IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens. , vol. 17, pp. 2629–2639, 2024

    work page 2024

  2. [49]

    A unified deep learning network for remote sensing image registration and change detection,

    R. Zhou, D. Quan, S. Wang, C. Lv, X. Cao, J. Chanussot, Y . Li, and L. Jiao, “A unified deep learning network for remote sensing image registration and change detection,” IEEE Trans. Geosci. Remote Sens. , vol. 62, pp. 1–16, 2024, Art. no. 5101216

    work page 2024

  3. [1]

    A review of change detection in multitemporal hyperspectral images: Current techniques, applications, and challenges,

    S. Liu, D. Marinelli, L. Bruzzone, and F. Bovolo, “A review of change detection in multitemporal hyperspectral images: Current techniques, applications, and challenges,” IEEE Geosci. Remote Sens. Mag. , vol. 7, no. 2, pp. 140–158, 2019

    work page 2019

  4. [2]

    An automatic change detection method for monitoring newly constructed building areas using time-series multi-view high-resolution optical satellite images,

    X. Huang, Y . Cao, and J. Li, “An automatic change detection method for monitoring newly constructed building areas using time-series multi-view high-resolution optical satellite images,” Remote Sens. Environ. , vol. 244, pp. 111 802–1–18, 2020

    work page 2020

  5. [3]

    Deep learning-based change detection in remote sensing images: A review,

    A. Shafique, G. Cao, Z. Khan, M. Asad, and M. Aslam, “Deep learning-based change detection in remote sensing images: A review,” Remote Sen. , vol. 14, no. 4, p. 871, 2022

    work page 2022

  6. [4]

    A review of data augmentation methods of remote sensing image target recognition,

    X. Hao, L. Liu, R. Yang, L. Yin, L. Zhang, and X. Li, “A review of data augmentation methods of remote sensing image target recognition,” Remote Sens. , vol. 15, no. 3, p. 827, 2023

    work page 2023

  7. [5]

    Fourier domain structural relationship analysis for unsupervised multimodal change de- tection,

    H. Chen, N. Yokoya, and M. Chini, “Fourier domain structural relationship analysis for unsupervised multimodal change de- tection,” ISPRS J. Photogramm. Remote Sens. , vol. 198, pp. 99–114, 2023

    work page 2023

  8. [6]

    Self-supervised pre- training via multimodality images with transformer for change detection,

    Y . Zhang, Y . Zhao, Y . Dong, and B. Du, “Self-supervised pre- training via multimodality images with transformer for change detection,” IEEE Trans. Geosci. Remote Sens., vol. 61, pp. 1–11, 2023, Art. no. 5402711

    work page 2023

Show all 65 references
  1. [7]

    Ssl4eo-s12: A large-scale multimodal, multi- temporal dataset for self-supervised learning in earth observa- tion [software and data sets],

    Y . Wang, N. A. A. Braham, Z. Xiong, C. Liu, C. M. Albrecht, and X. X. Zhu, “Ssl4eo-s12: A large-scale multimodal, multi- temporal dataset for self-supervised learning in earth observa- tion [software and data sets],” IEEE Geosci. Remote Sens. Mag. , vol. 11, no. 3, pp. 98–106, 2023

    2023

  2. [8]

    Exploring fine- grained image-text alignment for referring remote sensing image segmentation,

    S. Lei, X. Xiao, H.-C. Li, Z. Shi, and Q. Zhu, “Exploring fine- grained image-text alignment for referring remote sensing image segmentation,” IEEE Trans. Geosci. Remote Sens. , 2024, Early Access

    2024

  3. [9]

    Ss-mae: Spa- tial–spectral masked autoencoder for multisource remote sens- ing image classification,

    J. Lin, F. Gao, X. Shi, J. Dong, and Q. Du, “Ss-mae: Spa- tial–spectral masked autoencoder for multisource remote sens- ing image classification,” IEEE Trans. Geosci. Remote Sens. , vol. 61, pp. 1–14, 2023, Art. no. 5531614

    2023

  4. [10]

    Snunet-cd: A densely connected siamese network for change detection of vhr images,

    S. Fang, K. Li, J. Shao, and Z. Li, “Snunet-cd: A densely connected siamese network for change detection of vhr images,” IEEE Geosci. Remote Sens. Lett. , vol. 19, pp. 1–5, 2021, Art. no. 8007805

    2021

  5. [11]

    A feature difference convolutional neural network-based change detection method,

    M. Zhang and W. Shi, “A feature difference convolutional neural network-based change detection method,” IEEE Trans. Geosci. Remote Sens. , vol. 58, no. 10, pp. 7232–7246, 2020

    2020

  6. [12]

    An attention- based multiscale transformer network for remote sensing image change detection,

    W. Liu, Y . Lin, W. Liu, Y . Yu, and J. Li, “An attention- based multiscale transformer network for remote sensing image change detection,” ISPRS J. Photogramm. Remote Sens. , vol. 202, pp. 599–609, 2023

    2023

  7. [13]

    A full-level fused cross-task transfer learning method for building change detection using noise- robust pretrained networks on crowdsourced labels,

    Y . Cao and X. Huang, “A full-level fused cross-task transfer learning method for building change detection using noise- robust pretrained networks on crowdsourced labels,” Remote Sens. Environ., vol. 284, pp. 113 371–1–23, 2023

    2023

  8. [14]

    Sea ice change detection in sar images based on convolutional-wavelet neural networks,

    F. Gao, X. Wang, Y . Gao, J. Dong, and S. Wang, “Sea ice change detection in sar images based on convolutional-wavelet neural networks,” IEEE Geosci. Remote Sens. Lett. , vol. 16, no. 8, pp. 1240–1244, 2019

    2019

  9. [15]

    Remote sensing image change detection with transformers,

    H. Chen, Z. Qi, and Z. Shi, “Remote sensing image change detection with transformers,” IEEE Trans. Geosci. Remote Sens., vol. 60, pp. 1–14, 2021, Art. no. 5607514

    2021

  10. [16]

    Dam- net: Flood detection from sar imagery using differential atten- tion metric-based vision transformers,

    T. Saleh, X. Weng, S. Holail, C. Hao, and G.-S. Xia, “Dam- net: Flood detection from sar imagery using differential atten- tion metric-based vision transformers,” ISPRS J. Photogramm. Remote Sens. , vol. 212, pp. 440–453, 2024

    2024

  11. [17]

    Global and local attention-based transformer for hyperspectral image change detection,

    Z. Wang, F. Gao, J. Dong, and Q. Du, “Global and local attention-based transformer for hyperspectral image change detection,” IEEE Geosci. Remote Sens. Lett. , pp. 1–1, 2024

    2024

  12. [18]

    Ida-siamnet: Interactive-and dynamic-aware siamese network for building change detection,

    Y .-C. Li, S. Lei, N. Liu, H.-C. Li, and Q. Du, “Ida-siamnet: Interactive-and dynamic-aware siamese network for building change detection,” IEEE Trans. Geosci. Remote Sens. , 2024, Art. no. 5628213

    2024

  13. [19]

    A novel remote sensing image change detection approach based on multi-level state space model,

    Z. Zhang, X. Fan, X. Wang, Y . Qin, and J. Xia, “A novel remote sensing image change detection approach based on multi-level state space model,” IEEE Trans. Geosci. Remote Sens. , vol. 62, pp. 1–14, 2024, Art. no. 4417014

    2024

  14. [20]

    Change- mamba: Remote sensing change detection with spatio-temporal state space model,

    H. Chen, J. Song, C. Han, J. Xia, and N. Yokoya, “Change- mamba: Remote sensing change detection with spatio-temporal state space model,” IEEE Trans. Geosci. Remote Sens. , vol. 62, pp. 1–20, 2024, Art. no. 4409720

    2024

  15. [21]

    Cd- mamba: Remote sensing image change detection with mamba,

    H. Zhang, K. Chen, C. Liu, H. Chen, Z. Zou, and Z. Shi, “Cd- mamba: Remote sensing image change detection with mamba,” arXiv preprint arXiv:2406.04207 , 2024

    2024 arXiv

  16. [22]

    Unifying remote sensing change detection via deep probabilistic change models: From principles, models to applications,

    Z. Zheng, Y . Zhong, J. Zhao, A. Ma, and L. Zhang, “Unifying remote sensing change detection via deep probabilistic change models: From principles, models to applications,” ISPRS J. Photogramm. Remote Sens. , vol. 215, pp. 239–255, 2024

    2024

  17. [23]

    A mamba-based siamese network for remote sensing change detection,

    J. N. Paranjape, C. de Melo, and V . M. Patel, “A mamba-based siamese network for remote sensing change detection,” arXiv preprint arXiv:2407.06839, 2024

    2024 arXiv

  18. [24]

    Dc- mamba: A novel network for enhanced remote sensing change detection in difficult cases,

    J. Zhang, R. Chen, F. Liu, H. Liu, B. Zheng, and C. Hu, “Dc- mamba: A novel network for enhanced remote sensing change detection in difficult cases,” Remote Sens. , vol. 16, no. 22, p. 4186, 2024

    2024

  19. [25]

    A spatial-temporal attention-based method and a new dataset for remote sensing image change detection,

    H. Chen and Z. Shi, “A spatial-temporal attention-based method and a new dataset for remote sensing image change detection,” Remote Sens. , vol. 12, no. 10, p. 1662, 2020

    2020

  20. [26]

    Fully convolutional networks for multisource building extraction from an open aerial and satellite imagery data set,

    S. Ji, S. Wei, and M. Lu, “Fully convolutional networks for multisource building extraction from an open aerial and satellite imagery data set,” IEEE Trans. Geosci. Remote Sens. , vol. 57, no. 1, pp. 574–586, 2019

    2019

  21. [27]

    Change detection in remote sensing images using conditional adversarial networks,

    M. Lebedev, Y . V . Vizilter, O. Vygolov, V . A. Knyaz, and A. Y . Rubis, “Change detection in remote sensing images using conditional adversarial networks,” Int. Arch. Photogramm. 15 Remote Sens. Spatial Inf. Sci. , vol. 42, pp. 565–571, 2018

    2018

  22. [28]

    Loftr: Detector-free local feature matching with transformers,

    J. Sun, Z. Shen, Y . Wang, H. Bao, and X. Zhou, “Loftr: Detector-free local feature matching with transformers,” inProc. IEEE Conf. Comput. Vis. Pattern Recognit. , 2021, pp. 8922– 8931

    2021

  23. [29]

    Superglue: Learning feature matching with graph neural net- works,

    P.-E. Sarlin, D. DeTone, T. Malisiewicz, and A. Rabinovich, “Superglue: Learning feature matching with graph neural net- works,” in Proc. IEEE Conf. Comput. Vis. Pattern Recognit. , 2020, pp. 4938–4947

    2020

  24. [30]

    Lightglue: Local feature matching at light speed,

    P. Lindenberger, P.-E. Sarlin, and M. Pollefeys, “Lightglue: Local feature matching at light speed,” in Proc. IEEE Int. Conf. Comput. Vis., 2023, pp. 17 627–17 638

    2023

  25. [31]

    Dkm: Dense kernelized feature matching for geometry esti- mation,

    J. Edstedt, I. Athanasiadis, M. Wadenb ¨ack, and M. Felsberg, “Dkm: Dense kernelized feature matching for geometry esti- mation,” in Proc. IEEE Conf. Comput. Vis. Pattern Recognit. , 2023, pp. 17 765–17 775

    2023

  26. [32]

    Toward distortion- aware change detection in realistic scenarios,

    Y . Zhao, H.-C. Li, N. Liu, and R. Wang, “Toward distortion- aware change detection in realistic scenarios,” in IEEE Int. Geosci. Remote Sens. Symp. , 2024, pp. 10 319–10 322

    2024

  27. [33]

    Segment anything,

    A. Kirillov, E. Mintun, N. Ravi, H. Mao, C. Rolland, L. Gustafson, T. Xiao, S. Whitehead, A. C. Berg, W.-Y . Lo et al. , “Segment anything,” in Proc. IEEE Int. Conf. Comput. Vis., 2023, pp. 4015–4026

    2023

  28. [34]

    Learning transferable visual models from natural language supervision,

    A. Radford, J. W. Kim, C. Hallacy, A. Ramesh, G. Goh, S. Agarwal, G. Sastry, A. Askell, P. Mishkin, J. Clark et al. , “Learning transferable visual models from natural language supervision,” in Proc. Int. Conf. Mach. Learn. PMLR, 2021, pp. 8748–8763

    2021

  29. [35]

    Transformer-based multistage enhancement for remote sensing image super-resolution,

    S. Lei, Z. Shi, and W. Mo, “Transformer-based multistage enhancement for remote sensing image super-resolution,” IEEE Trans. Geosci. Remote Sens. , vol. 60, pp. 1–11, 2021

    2021

  30. [36]

    Spectralgpt: Spectral remote sensing foundation model,

    D. Hong, B. Zhang, X. Li, Y . Li, C. Li, J. Yao, N. Yokoya, H. Li, P. Ghamisi, X. Jia, A. Plaza, P. Gamba, J. A. Benediktsson, and J. Chanussot, “Spectralgpt: Spectral remote sensing foundation model,” IEEE Trans. Pattern Anal. Mach. Intell. , vol. 46, no. 8, pp. 5227–5244, 2024

    2024

  31. [37]

    Pointsam: Pointly-supervised segment anything model for remote sensing images,

    N. Liu, X. Xu, Y . Su, H. Zhang, and H.-C. Li, “Pointsam: Pointly-supervised segment anything model for remote sensing images,” arXiv preprint arXiv:2409.13401 , 2024

    2024 arXiv

  32. [38]

    Sam-assisted remote sensing imagery semantic segmentation with object and boundary constraints,

    X. Ma, Q. Wu, X. Zhao, X. Zhang, M.-O. Pun, and B. Huang, “Sam-assisted remote sensing imagery semantic segmentation with object and boundary constraints,” IEEE Trans. Geosci. Remote Sens. , vol. 62, pp. 1–16, 2024, Art. no. 5636916

    2024

  33. [39]

    Rsprompter: Learning to prompt for remote sensing instance segmentation based on visual foundation model,

    K. Chen, C. Liu, H. Chen, H. Zhang, W. Li, Z. Zou, and Z. Shi, “Rsprompter: Learning to prompt for remote sensing instance segmentation based on visual foundation model,” IEEE Trans. Geosci. Remote Sens., vol. 62, pp. 1–17, 2024, Art. no. 4701117

    2024

  34. [40]

    Ringmo: A remote sensing foundation model with masked image modeling,

    X. Sun, P. Wang, W. Lu, Z. Zhu, X. Lu, Q. He, J. Li, X. Rong, Z. Yang, H. Chang et al., “Ringmo: A remote sensing foundation model with masked image modeling,” IEEE Trans. Geosci. Remote Sens. , vol. 61, pp. 1–22, 2022

    2022

  35. [41]

    Changeclip: Remote sensing change detection with multimodal vision-language rep- resentation learning,

    S. Dong, L. Wang, B. Du, and X. Meng, “Changeclip: Remote sensing change detection with multimodal vision-language rep- resentation learning,” ISPRS J. Photogramm. Remote Sens. , vol. 208, pp. 53–69, 2024

    2024

  36. [42]

    Geochat: Grounded large vision-language model for remote sensing,

    K. Kuckreja, M. S. Danish, M. Naseer, A. Das, S. Khan, and F. S. Khan, “Geochat: Grounded large vision-language model for remote sensing,” in Proc. IEEE Conf. Comput. Vis. Pattern Recognit., 2024, pp. 27 831–27 840

    2024

  37. [43]

    Distinctive image features from scale-invariant keypoints,

    D. G. Lowe, “Distinctive image features from scale-invariant keypoints,” Int. J. Comput. Vis. , vol. 60, pp. 91–110, 2004

    2004

  38. [44]

    Superpoint: Self-supervised interest point detection and description,

    D. DeTone, T. Malisiewicz, and A. Rabinovich, “Superpoint: Self-supervised interest point detection and description,” in Proc. IEEE Conf. Comput. Vis. Pattern Recognit. Worksh., 2018, pp. 224–236

    2018

  39. [45]

    Gim: Learning generalizable image matcher from internet videos,

    X. Shen, Z. Cai, W. Yin, M. M ¨uller, Z. Li, K. Wang, X. Chen, and C. Wang, “Gim: Learning generalizable image matcher from internet videos,” in Proc. Int. Conf. Learn. Represent. , 2024

    2024

  40. [46]

    Geo- metric and non-linear radiometric distortion robust multimodal image matching via exploiting deep feature maps,

    M. Chen, Y . Zhao, T. Fang, Q. Zhu, S. Yan, and F. Gao, “Geo- metric and non-linear radiometric distortion robust multimodal image matching via exploiting deep feature maps,” ISPRS Ann. Photogramm. Remote Sens. Spatial Inf. Sci. , vol. 3, pp. 233– 240, 2020

    2020

  41. [47]

    Robust matching for sar and optical images using multiscale convolutional gra- dient features,

    L. Zhou, Y . Ye, T. Tang, K. Nan, and Y . Qin, “Robust matching for sar and optical images using multiscale convolutional gra- dient features,” IEEE Geosci. Remote Sens. Lett. , vol. 19, pp. 1–5, 2022, Art. no. 4017605

    2022

  42. [50]

    Momentum contrast for unsupervised visual representation learning,

    K. He, H. Fan, Y . Wu, S. Xie, and R. Girshick, “Momentum contrast for unsupervised visual representation learning,” in Proc. IEEE Conf. Comput. Vis. Pattern Recognit. , 2020, pp. 9729–9738

    2020

  43. [51]

    A simple framework for contrastive learning of visual representations,

    T. Chen, S. Kornblith, M. Norouzi, and G. Hinton, “A simple framework for contrastive learning of visual representations,” in Proc. Int. Conf. Mach. Learn. , 2020, pp. 1597–1607

    2020

  44. [52]

    Exploring simple siamese representation learning,

    X. Chen and K. He, “Exploring simple siamese representation learning,” in Proc. IEEE Conf. Comput. Vis. Pattern Recognit. , 2021, pp. 15 750–15 758

    2021

  45. [53]

    Masked autoencoders are scalable vision learners,

    K. He, X. Chen, S. Xie, Y . Li, P. Doll ´ar, and R. Girshick, “Masked autoencoders are scalable vision learners,” in Proc. IEEE Conf. Comput. Vis. Pattern Recognit. , 2022, pp. 16 000– 16 009

    2022

  46. [54]

    Self-supervised global–local contrastive learning for fine- grained change detection in vhr images,

    F. Jiang, M. Gong, H. Zheng, T. Liu, M. Zhang, and J. Liu, “Self-supervised global–local contrastive learning for fine- grained change detection in vhr images,” IEEE Trans. Geosci. Remote Sens. , vol. 61, pp. 1–13, 2023, Art. no. 4400613

    2023

  47. [55]

    Sslchange: A self-supervised change detection framework based on domain adaptation,

    Y . Zhao, T. Celik, N. Liu, F. Gao, and H.-C. Li, “Sslchange: A self-supervised change detection framework based on domain adaptation,” IEEE Trans. Geosci. Remote Sens. , vol. 62, pp. 1– 14, 2024, Art. no. 5647814

    2024

  48. [56]

    Coud: Continual urbanization detector for time series building change detection,

    Y . Zhao, H.-C. Li, S. Lei, N. Liu, J. Pan, and T. Celik, “Coud: Continual urbanization detector for time series building change detection,” IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens. , vol. 17, pp. 19 601–19 615, 2024

    2024

  49. [57]

    Fast segment anything,

    X. Zhao, W. Ding, Y . An, Y . Du, T. Yu, M. Li, M. Tang, and J. Wang, “Fast segment anything,” arXiv preprint arXiv:2306.12156, 2023

    2023 arXiv

  50. [58]

    Emerging properties in self-supervised vision transformers,

    M. Caron, H. Touvron, I. Misra, H. J ´egou, J. Mairal, P. Bo- janowski, and A. Joulin, “Emerging properties in self-supervised vision transformers,” in Proc. IEEE Int. Conf. Comput. Vis. , 2021, pp. 9650–9660

    2021

  51. [59]

    A computer algorithm for reconstruct- ing a scene from two projections,

    H. C. Longuet-Higgins, “A computer algorithm for reconstruct- ing a scene from two projections,” Nature, vol. 293, no. 5828, pp. 133–135, 1981

    1981

  52. [60]

    Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography,

    M. A. Fischler and R. C. Bolles, “Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography,” Commun. ACM, vol. 24, no. 6, pp. 381–395, 1981

    1981

  53. [61]

    Aid: A benchmark data set for performance evaluation of aerial scene classification,

    G.-S. Xia, J. Hu, F. Hu, B. Shi, X. Bai, Y . Zhong, L. Zhang, and X. Lu, “Aid: A benchmark data set for performance evaluation of aerial scene classification,” IEEE Trans. Geosci. Remote Sens., vol. 55, no. 7, pp. 3965–3981, 2017

    2017

  54. [62]

    Bag-of-visual-words and spatial extensions for land-use classification,

    Y . Yang and S. Newsam, “Bag-of-visual-words and spatial extensions for land-use classification,” inSIGSPATIAL Int. Conf. Adv. Geographic Inf. Syst. , 2010, pp. 270–279

    2010

  55. [63]

    Fully convolutional siamese networks for change detection,

    R. C. Daudt, B. Le Saux, and A. Boulch, “Fully convolutional siamese networks for change detection,” in Proc. IEEE Int. Conf. Image Process. , 2018, pp. 4063–4067

    2018

  56. [64]

    Ultralightweight spatial–spectral feature cooperation network for change detection in remote sensing images,

    T. Lei, X. Geng, H. Ning, Z. Lv, M. Gong, Y . Jin, and A. K. 16 Nandi, “Ultralightweight spatial–spectral feature cooperation network for change detection in remote sensing images,” IEEE Trans. Geosci. Remote Sens. , vol. 61, pp. 1–14, 2023, Art. no. 4402114

    2023

  57. [65]

    Spatiotemporal enhance- ment and interlevel fusion network for remote sensing images change detection,

    Y . Huang, X. Li, Z. Du, and H. Shen, “Spatiotemporal enhance- ment and interlevel fusion network for remote sensing images change detection,” IEEE Trans. Geosci. Remote Sens. , vol. 62, pp. 1–14, 2024, Art. no. 5609414

    2024

Pith tools

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