Pith. sign in

REVIEW 3 major objections 5 minor 67 references

Detect Changes like Humans: Incorporating Semantic Priors for Improved Change Detection

T0 review · 3 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read Change detectors improve when the network also knows what the landscape contains.

desk verdict Solid incremental change-detection work with a useful pretraining recipe, but the SOTA claim needs qualification and the WHU-CD pretraining gain needs an overlap audit. read the letter →

arxiv 2412.16918 v2 pith:VNQXYKXZ submitted 2024-12-22 cs.CV

Yuhang Gan , Wenjie Xuan , Zhiming Luo , Lei Fang , Zengmao Wang , Juhua Liu , Bo Du This is my paper · ORCID
classification cs.CV
keywords changedetectionremotesensingvisualfoundationmodelsemanticpriorsdual-streamdecodersingle-temporalpretrainingpseudo-changedataFastSAM
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper argues that binary change detection in remote sensing is held back by training on binary change maps alone, which teaches networks to spot appearance differences but not to understand what the landscape is. It proposes SA-CDNet, a network that keeps a frozen visual foundation model (FastSAM) as encoder, aligns its features with a lightweight adapter, and decodes semantic-aware and difference-aware features in two separate branches before fusing them with a learned weight. To supply semantic knowledge that small change-detection datasets lack, the paper pretrains the adapter and decoder on pseudo-change data built by pairing single-temporal segmentation images and labeling changes as the XOR of their segmentation maps. On five building- and farmland-change benchmarks the pretrained network reports the best F1 scores, with larger gains on datasets that contain seasonal, illumination, or fine-grained changes. If correct, this shows that semantics from segmentation data and frozen foundation models can be transferred into change detection without requiring paired bi-temporal change annotations.

What carries the argument

The key machinery is the dual-stream decoder plus the pseudo-change pretraining objective. The semantic-aware branch decodes each temporal image's features separately and fuses them late, keeping object semantics intact; the multi-scale difference-aware branch uses densely connected decoding units that concatenate the two images' features to capture change at multiple scales. A learnable weight $\sigma(\omega)$ fuses the two branches' predictions into the final map. The pretraining stage constructs pseudo-bi-temporal pairs from single-temporal segmentation images, sets the change label to the XOR of the two segmentation maps, and appends a segmentation head that imposes a proxy semantic segmentation loss on each image, so the adapter and decoder are initialized with landscape semantics before fine-tuning on real change maps.

What would settle it

Train SA-CDNet with the same pretraining pipeline but replace the XOR-derived change label with a random binary map while keeping the extra segmentation head; if downstream F1 gains on WHU-CD persist, then the pseudo-change detection loss and the proxy assumption behind it are not what carries the benefit.

Watch

Extended reading notes

Core claim

Bi-temporal supervision alone, the paper claims, pushes networks toward difference-aware features and away from object semantics, which is why they are brittle under noise and illumination variation. The central discovery is that a change detector profits from explicitly maintaining both feature types: a semantic-aware decoder that processes each temporal image independently and fuses late, preserving per-image object structure, and a multi-scale difference-aware decoder that concatenates the two images' features to emphasize change; an adaptive fusion module with a learnable weight $\sigma(\omega)$ combines their predictions. The paper additionally shows that a single-temporal pretraining stage, where random pairs of segmentation images are treated as pseudo-bi-temporal pairs with the XOR of their semantic maps as the change label and an extra segmentation head supervises each image, improves fine-tuned performance and reports state-of-the-art F1 scores on LEVIR-CD, LEVIR-CD+, S2Looking, WHU-CD, and WHU Cultivated Land.

Load-bearing premise

The pretraining benefit rests on the assumption that pairing two unrelated single-temporal images and labeling the pixels where their object maps disagree teaches something useful about real changes between registered images of the same place.

Editorial extensions

If this is right

  • Change detection models can be pretrained without any bi-temporal change annotations, using only single-temporal segmentation data that are cheaper to collect.
  • Because the foundation-model encoder stays frozen, only the adapter, decoder branches, and fusion weight need to be trained, keeping the trainable parameter count small.
  • The two decoder branches are complementary: the ablations on WHU-CD show that adding the semantic-aware branch to the difference-aware branch raises F1, and a learnable fusion weight beats max or mean fusion.
  • Pretraining on segmentation datasets whose categories match the target change (building data for building change, farmland-filtered data for farmland change) helps more than mixing in unrelated landscape classes.
  • The recipe transfers to a non-building change category (farmland), suggesting the semantic-prior mechanism is not specific to buildings.

Reading between the lines

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

  • Beyond the paper: the same pseudo-change pretraining should transfer to other change categories such as roads, water, or vegetation whenever a single-temporal segmentation dataset with that class is available.
  • Beyond the paper: the single-class versus multi-class ablation suggests a testable representation-level explanation, namely that mixed-class pretraining may push the decoder to attend to class boundaries irrelevant to the target change, which could be checked by measuring feature separability before fine-tuning.
  • Beyond the paper: the semi-supervised results indicate the pretraining is a strong inductive bias, so a natural extension is to combine it with pseudo-labelling of unlabeled bi-temporal images rather than ignoring them as done here.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper proposes SA-CDNet, a binary change detection network that combines a frozen FastSAM encoder with a lightweight adapter, a dual-stream decoder (semantic-aware and difference-aware branches), and an adaptive fusion module. It further adopts a single-temporal semantic pre-training strategy in which pseudo-change pairs are built by randomly pairing images from single-temporal segmentation datasets and labeling changes as the XOR of the segmentation maps (Eq. 7); an auxiliary segmentation head is added during pre-training. The authors report experiments on five change detection benchmarks and claim state-of-the-art performance, along with extensive ablations on encoder choice, decoder components, fusion strategies, and pre-training dataset settings.

Significance. If the central claims held, the paper would provide useful evidence that semantic priors from a frozen visual foundation model and from single-temporal segmentation data can improve binary change detection beyond difference-only supervision. The manuscript has concrete strengths: the module ablations are systematic, the pseudo-change construction and pre-training pipeline are clearly described, the code is released, and the comparison includes several recent methods, many of them re-implemented by the authors. However, two load-bearing issues undermine the claims as stated: the reported F1 numbers do not support the 'state-of-the-art on all datasets' assertion, and the main pre-training evidence on WHU-CD is potentially inflated by geographic overlap between the pre-training corpus (WHU-Building) and the evaluation benchmark (WHU-CD).

major comments (3)
  1. [Sec. IV-C, Table X] The abstract, introduction, and Sec. IV-C claim that SA-CDNet achieves state-of-the-art performance on all five benchmarks, but Table X itself lists TTP [37] and SFCD-Net [38] with higher F1 on LEVIR-CD (92.26 vs. 91.53), LEVIR-CD+ (85.08 vs. 84.43), S2Looking (67.38 vs. 66.48), and WHU-Cul (75.27 vs. 75.20), and a comparable F1 on WHU-CD (94.37 vs. 94.47 for TTP). The caveat that the comparison excludes SAM-l-based methods appears only in the body text, not in the abstract or conclusion. The SOTA claim must be revised to specify the comparison regime, or the comparison should include these methods in the headline numbers. As written, the claim is contradicted by the paper's own table.
  2. [Sec. IV-B2 and Sec. IV-C, Tables VII, IX, XIII] The main evidence for the single-temporal semantic pre-training strategy rests on WHU-CD: Table XIII shows the largest gain there (+2.76 F1), and nearly all pre-training ablations in Sec. IV-B are conducted on WHU-CD. However, the pre-training corpus includes WHU-Building (Table I, ref. [53]) and the evaluation benchmark WHU-CD (Table II, also ref. [53]) originate from the same WHU dataset and cover the same geographic region. The paper explicitly calls WHU-CD 'unseen' in Sec. IV-B2, but if WHU-Building tiles overlap the WHU-CD area, the pre-training gains may reflect memorization of building layouts rather than transferable semantic priors. The authors should audit the spatial overlap between WHU-Building and WHU-CD, and re-run the WHU-CD ablations and the WHU-CD row of Tables X and XIII with a pre-training corpus that excludes geographically overlapping imagery. Without this audit, the central claim that single-temporal semantic pre-training improves change detection is not established for the benchmark on which it is primarily tested.
  3. [Sec. IV, Tables IV-XIII] All reported results are single-run point estimates, with no standard deviations or significance tests. Several decisive comparisons in the paper are decided by small margins: for example, WHU-Cul F1 is 75.20 vs. 75.27 for TTP in Table X, and some pre-training gains in Table XIII are as small as +0.64 F1. Without repeated runs, the claimed improvements from pre-training and the differential conclusions across pre-training datasets (e.g., Table VII) are not statistically supported. At minimum, the authors should run multiple seeds (e.g., three) for the main claims and report mean and standard deviation.
minor comments (5)
  1. [Sec. IV-C] In the sentence listing the new SOTA numbers, 'a75.20% F1-score on LEVIR-CD+' should read 'a75.20% F1-score on WHU-Cul'.
  2. [Sec. IV-C] The text says the pre-training 'surpass[es] the results without pre-training by a large margin', but the LEVIR-CD gain is +0.38 F1 (Table XIII), which is not a large margin; please qualify the statement accordingly.
  3. [Table X] The last row 'Ours (w pre-train)' reports the WHU-Cul F1 as 75.20 in the WHU-Cul column, but the text in Sec. IV-C attributes this number to LEVIR-CD+; please ensure the table and text are consistent.
  4. [Table III] The 'FastSAM-s' row lists '11.221.85' under Params/FLOPs; this appears to be a formatting issue that should be corrected to two separate numbers.
  5. [Sec. IV-B1] The phrase 'unseen WHU-CD dataset' is used to motivate that higher pre-training performance indicates better knowledge of remote sensing images; given the likely overlap between WHU-Building and WHU-CD, this characterization should be revised or justified with an overlap analysis.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the derivation relies on external data, external foundation models, and external prior work; self-citations are only baselines.

full rationale

SA-CDNet's derivation chain is self-contained: the frozen FastSAM encoder, adapter, dual-stream decoder, and fusion weights are trained on external segmentation and change-detection datasets; no equation defines its outputs in terms of its own predictions. Pseudo-change labels are external ground-truth segmentation masks combined by XOR (Eq. 7), and the proxy segmentation loss (Eq. 10) supervises with the same external masks; the fine-tuning loss (Eq. 12) uses external benchmark labels. The method adopts the pseudo-change construction from external work [15], and the SAM-CD-style adapter is externally grounded in [36]. Although the authors include their own RFL-CDNet [65] and group-related methods as baselines, these are comparison entries, not premises of the derivation, so they are not load-bearing. A separate data-source overlap between WHU-Building and WHU-CD is a potential test-set contamination/correctness risk, not a construction-level circularity, so it does not affect this score.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

The central claim rests on four domain assumptions and two tuned hyperparameters. No genuinely new entities are introduced. The most important assumption is that pseudo-change pairs from random segmentation images approximate real bi-temporal change; this comes from [15] and is load-bearing for the pretraining contribution.

free parameters (2)
  • Pre-training loss weight lambda = 1.0
    Eq. 11 sets the proxy segmentation loss weight lambda empirically to 1.0; the paper does not sweep it.
  • Adaptive fusion weight omega = learned, final value not reported
    Eq. 6 uses sigma(omega) to linearly combine semantic-aware and difference-aware predictions; the value is fit per dataset and no final value or sensitivity analysis is reported.
assumptions (4)
  • domain assumption FastSAM's frozen encoder features retain semantic priors useful for remote sensing change detection after a simple adapter.
    Invoked in Sec III-A1; the paper validates by comparing encoders and adapter ablations, but assumes transferability across the natural-to-remote-sensing domain gap.
  • domain assumption Pseudo-change pairs from random images with XOR segmentation labels approximate real bi-temporal changes.
    Sec III-B1 and Eq. 7 construct change labels as Yi XOR Yj; this proxy, adopted from [15], is load-bearing for the pretraining contribution.
  • domain assumption Training on binary cross-entropy with benchmark labels and fixed train/test splits gives unbiased evaluation.
    Sec IV-A assumes the public benchmark splits and labels are correct and representative for measuring generalization.
  • domain assumption WHU-CD is an unseen target for measuring pretraining transfer when WHU-Building segmentation data are used for pretraining.
    Sec IV-B2 and Tab. VII call WHU-CD unseen, but both WHU-Building and WHU-CD cite [53], so independence is not established.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Detect Changes like Humans: Incorporating Semantic Priors for Improved Change Detection." pith.science (2026). https://pith.science/paper/VNQXYKXZ

@misc{pith2026241216918,
  author       = {Pith},
  title        = {Pith review of: Detect Changes like Humans: Incorporating Semantic Priors for Improved Change Detection},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VNQXYKXZ}},
  note         = {Machine review of arXiv:2412.16918}
}
abstract

When given two similar images, humans identify their differences by comparing the appearance (e.g., color, texture) with the help of semantics (e.g., objects, relations). However, mainstream binary change detection models adopt a supervised training paradigm, where the annotated binary change map is the main constraint. Thus, such methods primarily emphasize difference-aware features between bi-temporal images, and the semantic understanding of changed landscapes is undermined, resulting in limited accuracy in the face of noise and illumination variations. To this end, this paper explores incorporating semantic priors from visual foundation models to improve the ability to detect changes. Firstly, we propose a Semantic-Aware Change Detection network (SA-CDNet), which transfers the knowledge of visual foundation models (i.e., FastSAM) to change detection. Inspired by the human visual paradigm, a novel dual-stream feature decoder is derived to distinguish changes by combining semantic-aware features and difference-aware features. Secondly, we explore a single-temporal pre-training strategy for better adaptation of visual foundation models. With pseudo-change data constructed from single-temporal segmentation datasets, we employ an extra branch of proxy semantic segmentation task for pre-training. We explore various settings like dataset combinations and landscape types, thus providing valuable insights. Experimental results on five challenging benchmarks demonstrate the superiority of our method over the existing state-of-the-art methods. The code is available at $\href{https://github.com/DREAMXFAR/SA-CDNet}{github}$.

Figures

Figures reproduced from arXiv: 2412.16918 by the authors.

Figure 1
Figure 1. To locate changes between images, humans not only [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. The overview of SA-CDNet. We employ the frozen FastSAM encoder to extract features, then refined by an adapter [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. The structure of dual-stream feature decoder, which [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: The pipeline of single-temporal semantic pre-training. We first construct pseudo-change data from single-temporal [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: Visualized comparison of our method with previous SOTA methods. [PITH_FULL_IMAGE:figures/full_fig_p011_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

67 extracted references · 60 canonical work pages

  1. [53]

    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 Transactions on Geoscience and Remote Sensing, 2018

    work page 2018

  2. [37]

    Time travelling pixels: Bitemporal features integration with foundation model for remote sensing image change detection,

    K. Chen, C. Liu, W. Li, Z. Liu, H. Chen, H. Zhang, Z. Zou, and Z. Shi, “Time travelling pixels: Bitemporal features integration with foundation model for remote sensing image change detection,” in2024-2024 IEEE In- ternational Geoscience and Remote Sensing Symposium, 2024, pp. 8581–8584

    work page 2024

  3. [38]

    Integrating sam with feature interaction for remote sens- ing change detection,

    D. Zhang, F. Wang, L. Ning, Z. Zhao, J. Gao, and X. Li, “Integrating sam with feature interaction for remote sens- ing change detection,”IEEE Transactions on Geoscience and Remote Sensing, 2024

    work page 2024

  4. [1]

    Unet-like remote sensing change detection: A review of current models and research directions,

    C. Wu, L. Zhang, B. Du, H. Chen, J. Wang, and H. Zhong, “Unet-like remote sensing change detection: A review of current models and research directions,”IEEE Geoscience and Remote Sensing Magazine, 2024

    work page 2024

  5. [2]

    Cast: A cascade spectral-aware transformer for hyperspectral image change detection,

    X. Zhang, S. Tian, G. Wang, X. Tang, J. Feng, and L. Jiao, “Cast: A cascade spectral-aware transformer for hyperspectral image change detection,”IEEE Transac- tions on Geoscience and Remote Sensing, 2023

    work page 2023

  6. [3]

    Oral: An observational learning paradigm for unsupervised hyperspectral change detection,

    G. Wang, X. Zhang, Z. Peng, S. Tian, T. Zhang, X. Tang, and L. Jiao, “Oral: An observational learning paradigm for unsupervised hyperspectral change detection,”IEEE Transactions on Circuits and Systems for Video Technol- ogy, 2025

    work page 2025

  7. [5]

    A deeply supervised image fusion network for change detection in high resolution 12 bi-temporal remote sensing images,

    C. Zhang, P. Yue, D. Tapete, L. Jiang, B. Shangguan, L. Huang, and G. Liu, “A deeply supervised image fusion network for change detection in high resolution 12 bi-temporal remote sensing images,”ISPRS Journal of Photogrammetry and Remote Sensing, 2020

    work page 2020

  8. [6]

    A transformer-based siamese network for change detection,

    W. G. C. Bandara and V . M. Patel, “A transformer-based siamese network for change detection,” in2022 IEEE In- ternational Geoscience and Remote Sensing Symposium, 2022

    work page 2022

Show all 67 references
  1. [8]

    Semantic-aware dense representation learning for remote sensing image change detection,

    H. Chen, W. Li, S. Chen, and Z. Shi, “Semantic-aware dense representation learning for remote sensing image change detection,”IEEE Transactions on Geoscience and Remote Sensing, 2022

    2022

  2. [9]

    Semisupervised change detection with feature-prediction alignment,

    X. Zhang, X. Huang, and J. Li, “Semisupervised change detection with feature-prediction alignment,” IEEE Transactions on Geoscience and Remote Sensing, 2023

    2023

  3. [10]

    Changes- aware transformer: Learning generalized changes repre- sentation,

    D. Wang, L. Jiao, J. Chen, S. Yang, and F. Liu, “Changes- aware transformer: Learning generalized changes repre- sentation,”ArXiv preprint arXiv:2309.13619, 2023

    2023 arXiv

  4. [11]

    An end-to-end supervised domain adaptation framework for cross-domain change detection,

    J. Liu, W. Xuan, Y . Gan, Y . Zhan, J. Liu, and B. Du, “An end-to-end supervised domain adaptation framework for cross-domain change detection,”Pattern Recognition, 2022

    2022

  5. [12]

    Mctnet: A multi- scale cnn-transformer network for change detection in optical remote sensing images,

    W. Li, L. Xue, X. Wang, and G. Li, “Mctnet: A multi- scale cnn-transformer network for change detection in optical remote sensing images,” in2023 26th Interna- tional Conference on Information Fusion, 2023, pp. 1–5

    2023

  6. [13]

    Semantic relation-aware difference representation learn- ing for change captioning,

    Y . Tu, T. Yao, L. Li, J. Lou, S. Gao, Z. Yu, and C. Yan, “Semantic relation-aware difference representation learn- ing for change captioning,” inFindings of the Association for Computational Linguistics-IJCNLP 2021, 2021, pp. 63–73

    2021

  7. [14]

    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

  8. [15]

    Change is everywhere: Single-temporal supervised object change detection in remote sensing imagery,

    Z. Zheng, A. Ma, L. Zhang, and Y . Zhong, “Change is everywhere: Single-temporal supervised object change detection in remote sensing imagery,” inProceedings of the IEEE/CVF International Conference on Computer Vision, 2021, pp. 15 193–15 202

    2021

  9. [16]

    Fully con- volutional siamese networks for change detection,

    R. C. Daudt, B. Le Saux, and A. Boulch, “Fully con- volutional siamese networks for change detection,” in 2018 25th IEEE International Conference on Image Processing, 2018, pp. 4063–4067

    2018

  10. [17]

    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 Geoscience and Remote Sensing Letters, 2021

    2021

  11. [18]

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

    A. Dosovitskiy, L. Beyer, A. Kolesnikovet al., “An image is worth 16x16 words: Transformers for image recognition at scale,” inInternational Conference on Learning Representations, 2021

    2021

  12. [19]

    Swinsunet: Pure transformer network for remote sensing image change detection,

    C. Zhang, L. Wang, S. Cheng, and Y . Li, “Swinsunet: Pure transformer network for remote sensing image change detection,”IEEE Transactions on Geoscience and Remote Sensing, 2022

    2022

  13. [20]

    Remote sensing image change detection with transformers,

    H. Chen, Z. Qi, and Z. Shi, “Remote sensing image change detection with transformers,”IEEE Transactions on Geoscience and Remote Sensing, 2021

    2021

  14. [21]

    Semantic change detec- tion with asymmetric siamese networks,

    K. Yang, G.-S. Xia, Z. Liu, B. Du, W. Yang, M. Pelillo, and L. Zhang, “Semantic change detec- tion with asymmetric siamese networks,”ArXiv preprint arXiv:2010.05687, 2020

    2010 arXiv

  15. [22]

    Scdnet: A novel convolutional network for semantic change detection in high resolution optical remote sens- ing imagery,

    D. Peng, L. Bruzzone, Y . Zhang, H. Guan, and P. He, “Scdnet: A novel convolutional network for semantic change detection in high resolution optical remote sens- ing imagery,”International Journal of Applied Earth Observation and Geoinformation, 2021

    2021

  16. [23]

    Multi-task learn- ing for semantic change detection on vhr remote sensing images,

    Y . Zhou, J. Zhu, L. Huo, and C. Huo, “Multi-task learn- ing for semantic change detection on vhr remote sensing images,” in2022-2022 IEEE International Geoscience and Remote Sensing Symposium, 2022, pp. 3247–3250

    2022

  17. [24]

    Bi-temporal semantic reasoning for the semantic change detection in hr remote sensing images,

    L. Ding, H. Guo, S. Liu, L. Mou, J. Zhang, and L. Bruz- zone, “Bi-temporal semantic reasoning for the semantic change detection in hr remote sensing images,”IEEE Transactions on Geoscience and Remote Sensing, 2022

    2022

  18. [25]

    On the opportuni- ties and risks of foundation models,

    R. Bommasani, Hudsonet al., “On the opportuni- ties and risks of foundation models,”ArXiv preprint arXiv:2108.07258, 2021

    2021 arXiv

  19. [26]

    Segment anything,

    A. Kirillov, E. Mintun, N. Ravi, H. Mao, C. Rolland, L. Gustafson, T. Xiao, S. Whitehead, A. C. Berg, W.- Y . Loet al., “Segment anything,” inProceedings of the IEEE/CVF International Conference on Computer Vision, 2023, pp. 4015–4026

    2023

  20. [27]

    Efficientsam: Leveraged masked image pretraining for efficient segment anything,

    Y . Xiong, B. Varadarajan, L. Wu, X. Xiang, F. Xiao, C. Zhu, X. Dai, D. Wang, F. Sun, F. Iandolaet al., “Efficientsam: Leveraged masked image pretraining for efficient segment anything,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2024,...

    2024

  21. [28]

    Faster segment anything: Towards lightweight sam for mobile applications,

    C. Zhang, D. Han, Y . Qiao, J. U. Kim, S.-H. Bae, S. Lee, and C. S. Hong, “Faster segment anything: Towards lightweight sam for mobile applications,”ArXiv preprint arXiv:2306.14289, 2023

    2023 arXiv

  22. [29]

    Laddering vision foundation model for remote sensing image change detection,

    Y . Liu and G. Zhou, “Laddering vision foundation model for remote sensing image change detection,”Journal of Applied Remote Sensing, 2024

    2024

  23. [30]

    Fine- grained high-resolution remote sensing image change detection by sam-unet change detection model,

    X. Zhao, Z. Wu, Y . Chen, W. Zhou, and M. Wei, “Fine- grained high-resolution remote sensing image change detection by sam-unet change detection model,”Remote Sensing, p. 3620, 2024

    2024

  24. [31]

    Ass-cd: Adapting segment anything model and swin-transformer for change detec- tion in remote sensing images,

    C. Wei, X. Wu, and B. Wang, “Ass-cd: Adapting segment anything model and swin-transformer for change detec- tion in remote sensing images,”Remote Sensing, 2025

    2025

  25. [32]

    Segment anything model guided semantic knowledge learning for remote sensing change detec- tion,

    Z. Sun, H. Song, K. Zhang, G. Dong, L. Liang, and Y . Zhao, “Segment anything model guided semantic knowledge learning for remote sensing change detec- tion,” in2024-2024 IEEE International Conference on Acoustics, Speech and Signal Processing, 2024, pp. 5830–5834

    2024

  26. [33]

    Ded-sam: Adapting segment anything model 2 for dual encoder- decoder change detection,

    J. Qiu, W. Liu, E. Li, L. Zhang, and X. Li, “Ded-sam: Adapting segment anything model 2 for dual encoder- decoder change detection,”IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sens- 13 ing, 2024

    2024

  27. [34]

    Combining sam with limited data for change detection in remote sensing,

    J. Gao, D. Zhang, F. Wang, L. Ning, Z. Zhao, and X. Li, “Combining sam with limited data for change detection in remote sensing,”IEEE Transactions on Geoscience and Remote Sensing, 2025

    2025

  28. [35]

    Sam-cd: Change detection in remote sens- ing using segment anything model,

    F. ALTam, T. Khalid, A. Mathew, A. Carnell, and R. Souissi, “Sam-cd: Change detection in remote sens- ing using segment anything model,” inNeurIPS 2023 Workshop on Tackling Climate Change with Machine Learning, 2023

    2023

  29. [36]

    Adapting segment anything model for change detection in vhr remote sensing images,

    L. Ding, K. Zhu, D. Peng, H. Tang, K. Yang, and L. Bruzzone, “Adapting segment anything model for change detection in vhr remote sensing images,”IEEE Transactions on Geoscience and Remote Sensing, 2024

    2024

  30. [39]

    Siamese-sam: Remote sensing image change detection with siamese structure segment anything model,

    G. Wei, Y . Miao, and Z. Wang, “Siamese-sam: Remote sensing image change detection with siamese structure segment anything model,”Applied Sciences, 2025

    2025

  31. [40]

    Fastsam-based change detection network for remote sensing images,

    X. Kong, J. Wang, J. Shen, Z. Ling, C. Jing, D. Zhang, and Z. Hu, “Fastsam-based change detection network for remote sensing images,” in2024 5th International Conference on Geology, Mapping and Remote Sensing, 2024, pp. 53–58

    2024

  32. [41]

    Sam-based building change detection with distribution- aware fourier adaptation and edge-constrained warping,

    Y .-C. Li, S. Lei, Y .-T. Zhao, H.-C. Li, J. Li, and A. Plaza, “Sam-based building change detection with distribution- aware fourier adaptation and edge-constrained warping,” ArXiv preprint arXiv:2504.12619, 2025

    2025 arXiv

  33. [42]

    Feature pyramid networks for object detection,

    T.-Y . Lin, P. Dollar, R. Girshick, K. He, B. Hariharan, and S. Belongie, “Feature pyramid networks for object detection,” inProceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 2017

    2017

  34. [43]

    Segment everything everywhere all at once,

    X. Zou, J. Yang, H. Zhang, F. Li, L. Li, J. Wang, L. Wang, J. Gao, and Y . J. Lee, “Segment everything everywhere all at once,” inAdvances in Neural Informa- tion Processing Systems, 2023, pp. 19 769–19 782

    2023

  35. [44]

    Medical sam 2: Segment medical images as video via segment anything model 2,

    J. Zhu, Y . Qi, and J. Wu, “Medical sam 2: Segment medical images as video via segment anything model 2,”ArXiv preprint arXiv:2408.00874, 2024

    2024 arXiv

  36. [45]

    End-to-end change detection for high resolution satellite images using im- proved unet++,

    D. Peng, Y . Zhang, and H. Guan, “End-to-end change detection for high resolution satellite images using im- proved unet++,”Remote Sensing, 2019

    2019

  37. [46]

    Eca-net: Efficient channel attention for deep convolu- tional neural networks,

    Q. Wang, B. Wu, P. Zhu, P. Li, W. Zuo, and Q. Hu, “Eca-net: Efficient channel attention for deep convolu- tional neural networks,” inProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2020, pp. 11 534–11 542

    2020

  38. [47]

    End-to-end object detection with transformers,

    N. Carion, F. Massa, G. Synnaeve, N. Usunier, A. Kir- illov, and S. Zagoruyko, “End-to-end object detection with transformers,” inEuropean Conference on Com- puter Vision, 2020, pp. 213–229

    2020

  39. [48]

    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,” inProceedings of the IEEE/CVF International Conference on Computer Vision, 2021, pp. 10 012–10 022

    2021

  40. [49]

    Change detec- tion based on supervised contrastive learning for high- resolution remote sensing imagery,

    J. Wang, Y . Zhong, and L. Zhang, “Change detec- tion based on supervised contrastive learning for high- resolution remote sensing imagery,”IEEE Transactions on Geoscience and Remote Sensing, 2023

    2023

  41. [50]

    Unified building change detection pre- training method with masked semantic annotations,

    Y . Quan, A. Yu, W. Guo, X. Lu, B. Jiang, S. Zheng, and P. He, “Unified building change detection pre- training method with masked semantic annotations,”In- ternational Journal of Applied Earth Observation and Geoinformation, 2023

    2023

  42. [51]

    Temporary removal: Aerial imagery for roof segmentation: A large-scale dataset towards automatic mapping of buildings,

    Q. Chen, L. Wang, Y . Wu, G. Wu, Z. Guo, and S. L. Waslander, “Temporary removal: Aerial imagery for roof segmentation: A large-scale dataset towards automatic mapping of buildings,”ISPRS Journal of Photogramme- try and Remote Sensing, 2019

    2019

  43. [52]

    Can semantic labeling methods generalize to any city? the inria aerial image labeling benchmark,

    E. Maggiori, Y . Tarabalka, G. Charpiat, and P. Alliez, “Can semantic labeling methods generalize to any city? the inria aerial image labeling benchmark,” in2017 IEEE International Geoscience and Remote Sensing Sympo- sium, 2017, pp. 3226–3229

    2017

  44. [54]

    Deepglobe 2018: A challenge to parse the earth through satellite images,

    I. Demir, K. Koperski, D. Lindenbaum, G. Pang, J. Huang, S. Basu, F. Hughes, D. Tuia, and R. Raskar, “Deepglobe 2018: A challenge to parse the earth through satellite images,” inProceedings of the IEEE Conference on Computer Vision and Pattern Recognition workshops, 2018, pp. 172–181

    2018

  45. [55]

    Semantic segmentation of remote- sensing imagery using heterogeneous big data: Interna- tional society for photogrammetry and remote sensing potsdam and cityscape datasets,

    A. Song and Y . Kim, “Semantic segmentation of remote- sensing imagery using heterogeneous big data: Interna- tional society for photogrammetry and remote sensing potsdam and cityscape datasets,”ISPRS International Journal of Geo-Information, 2020

    2020

  46. [56]

    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 Sensing, 2020

    2020

  47. [57]

    S2looking: A satellite side-looking dataset for building change detection,

    L. Shen, Y . Lu, H. Chen, H. Wei, D. Xie, J. Yue, R. Chen, S. Lv, and B. Jiang, “S2looking: A satellite side-looking dataset for building change detection,”Remote Sensing, 2021

    2021

  48. [58]

    Whu cultivated land dataset,

    “Whu cultivated land dataset,” http://rsipac.whu.edu.cn/ subject two 2021

    2021

  49. [59]

    Stanet: A spatiotemporal attention network for decoding audi- tory spatial attention from eeg,

    E. Su, S. Cai, L. Xie, H. Li, and T. Schultz, “Stanet: A spatiotemporal attention network for decoding audi- tory spatial attention from eeg,”IEEE Transactions on Biomedical Engineering, 2022

    2022

  50. [60]

    Dasnet: Dual attentive fully convolu- tional siamese networks for change detection in high- resolution satellite images,

    J. Chen, Z. Yuan, J. Peng, L. Chen, H. Huang, J. Zhu, Y . Liu, and H. Li, “Dasnet: Dual attentive fully convolu- tional siamese networks for change detection in high- resolution satellite images,”IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sens- 14...

    2020

  51. [61]

    Remote sensing change detection based on multidirectional adap- tive feature fusion and perceptual similarity,

    J. Xu, C. Luo, X. Chen, S. Wei, and Y . Luo, “Remote sensing change detection based on multidirectional adap- tive feature fusion and perceptual similarity,”Remote Sensing, 2021

    2021

  52. [62]

    Fully transformer network for change detection of remote sensing images,

    T. Yan, Z. Wan, and P. Zhang, “Fully transformer network for change detection of remote sensing images,” inPro- ceedings of the Asian Conference on Computer Vision, 2022, pp. 1691–1708

    2022

  53. [63]

    Vct: Visual change transformer for remote sensing image change detection,

    B. Jiang, Z. Wang, X. Wang, Z. Zhang, L. Chen, X. Wang, and B. Luo, “Vct: Visual change transformer for remote sensing image change detection,”IEEE Trans- actions on Geoscience and Remote Sensing, 2023

    2023

  54. [64]

    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 Journal of Photogram- metry and Remote Sensing, 2023

    2023

  55. [65]

    Rfl-cdnet: Towards accurate change detection via richer feature learning,

    Y . Gan, W. Xuan, H. Chen, J. Liu, and B. Du, “Rfl-cdnet: Towards accurate change detection via richer feature learning,”Pattern Recognition, 2024

    2024

  56. [66]

    Semicdnet: A semisupervised convo- lutional neural network for change detection in high resolution remote-sensing images,

    D. Peng, L. Bruzzone, Y . Zhang, H. Guan, H. Ding, and X. Huang, “Semicdnet: A semisupervised convo- lutional neural network for change detection in high resolution remote-sensing images,”IEEE Transactions on Geoscience and Remote Sensing, 2020

    2020

  57. [67]

    Semisanet: A semi-supervised high-resolution remote sensing image change detection model using siamese networks with graph attention,

    C. Sun, J. Wu, H. Chen, and C. Du, “Semisanet: A semi-supervised high-resolution remote sensing image change detection model using siamese networks with graph attention,”Remote Sensing, 2022

    2022

  58. [68]

    Mtcnet: Multitask consistency network with single temporal su- pervision for semi-supervised building change detection,

    Q. Shu, J. Pan, Z. Zhang, and M. Wang, “Mtcnet: Multitask consistency network with single temporal su- pervision for semi-supervised building change detection,” International Journal of Applied Earth Observation and Geoinformation, 2022

    2022

  59. [69]

    Semicd-vl: Visual-language model guidance makes better semi-supervised change detector,

    K. Li, X. Cao, Y . Deng, J. Song, J. Liu, D. Meng, and Z. Wang, “Semicd-vl: Visual-language model guidance makes better semi-supervised change detector,”IEEE Transactions on Geoscience and Remote Sensing, 2025. Yuhang Ganis currently pursuing his Ph.D. de- gree in Electronic I...

    2025

Pith tools

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