The reviewed record of science sign in
Pith

arxiv: 2606.10329 · v1 · pith:MCK3PALR · submitted 2026-06-09 · cs.CV · cs.AI

Building Change Detection in Earthquake: A Multi-Scale Interaction Network and A Change Detection Dataset

Yunlong Liu , Zekai Zhang This is my paper

Reviewed by Pith T0 review T1 audit T2 compute T3 formal T4 kernel 2026-06-27 14:10 UTCgrok-4.3pith:MCK3PALRrecord.jsonopen to challenge →

classification cs.CV cs.AI
keywords change detectionremote sensingearthquake damagebuilding changemulti-scale networkoffset calibrationbi-temporal images
0
0 comments X

The pith

A multi-scale interaction network with offset calibration detects building changes in short-interval post-earthquake images.

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

The paper introduces the TUE-CD dataset of short-interval bi-temporal images from the Turkey earthquake to support immediate post-disaster building damage assessment. It proposes the MSI-Net architecture, which uses joint cross-attention modules to exchange information between image pairs, multi-scale offset calibration modules to align features distorted by differing viewing angles, and feature integration modules to combine the results for change maps. Experiments report that this network yields higher detection accuracy than prior methods on both existing benchmarks and the new TUE-CD collection. The work targets the practical constraint that standard change-detection models struggle when images are captured only days apart rather than months or years.

Core claim

MSI-Net, built from joint cross-attention, multi-scale offset calibration, and feature integration modules, produces more accurate building change maps than existing methods on the WHU-CD, CLCD, and newly collected TUE-CD datasets by explicitly estimating and correcting alignment offsets that arise from short-interval side-looking acquisitions.

What carries the argument

The multi-scale feature interaction network (MSI-Net) that unifies joint cross-attention for bi-temporal exchange, multi-scale offset calibration for alignment, and feature integration for final prediction.

If this is right

  • Short-interval post-event imagery becomes usable for rapid damage mapping instead of waiting for better-aligned acquisitions.
  • The offset calibration step reduces the side-looking artifacts that currently limit change detection after earthquakes.
  • The same three-module structure can be applied to other bi-temporal remote-sensing tasks that involve viewpoint shifts.
  • The TUE-CD dataset supplies a concrete benchmark for measuring progress on emergency-response change detection.

Where Pith is reading between the lines

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

  • The approach may transfer to other sudden-onset events such as floods or landslides where acquisition intervals are also short.
  • If the offset calibration proves robust, similar modules could be inserted into existing change-detection pipelines without retraining the entire network.
  • Longer-interval datasets could be used to test whether the side-looking correction introduces unnecessary complexity when viewing angles already match.

Load-bearing premise

Estimating offsets at multiple scales can align bi-temporal features from different imaging angles without creating new misalignment or noise.

What would settle it

If MSI-Net produces lower accuracy metrics than the strongest baseline methods when both are evaluated on the held-out TUE-CD test split, the performance advantage claim does not hold.

Figures

Figures reproduced from arXiv: 2606.10329 by Yunlong Liu, Zekai Zhang.

Figure 1
Figure 1. Figure 1: Mismatch problem appear in the TUE-CD, WHU-CD, CLCD [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Overall architecture of the proposed MSI-Net [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Architecture of JCA module [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Procedure of SJA block. bilinear interpolation is utilized to adjust the offset values in ∆P as integers for calibration. Then, the features Ei 1 and Ei 2 are calibrated by: Eˆi 1 (ˆp) = P q B (ˆp, q) Ei 1 (q) Eˆi 2 (ˆp) = P q B (ˆp, q) Ei 2 (q) (4) where pˆ = p + ∆p stands for the spatial position of sub￾pixels in features Ei 1 and Ei 2 . q is the integer sampling grid at the corresponding spatial locatio… view at source ↗
Figure 5
Figure 5. Figure 5: Architecture of MOC module. where Y denotes the ground truth and Yˆ is the predicted change result. w1 and w2 are set as 0.7 and 0.3, respectively. If the predicted change map value is 1, the pixel corresponding to that point is a change pixel, and vice versa. IV. EXPERIMENTS This section first describes two public datasets we used. Next, the constructed TUE-CD dataset for the assessment of building damage… view at source ↗
Figure 6
Figure 6. Figure 6: The most severely affected areas and some typical samples from these areas. [PITH_FULL_IMAGE:figures/full_fig_p008_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Qualitative comparison of all methods on the WHU-CD dataset. (a)–(e) Prediction results of all methods on different [PITH_FULL_IMAGE:figures/full_fig_p008_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Qualitative comparison of all methods on the CLCD dataset. (a)–(e) Prediction results of all methods on different [PITH_FULL_IMAGE:figures/full_fig_p009_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Qualitative comparison of all methods on the TUE-CD dataset. (a)–(e) Prediction results of all methods on different [PITH_FULL_IMAGE:figures/full_fig_p009_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: Examples of qualitative results of the ablation study on the WHU-CD dataset. (a)–(e) Prediction results of all [PITH_FULL_IMAGE:figures/full_fig_p010_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: Examples of qualitative results of the ablation study on the CLCD dataset. (a)–(e) Prediction results of all methods [PITH_FULL_IMAGE:figures/full_fig_p011_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: Examples of qualitative results of the ablation study on the TUE-CD dataset. (a)–(e) Prediction results of all [PITH_FULL_IMAGE:figures/full_fig_p012_12.png] view at source ↗
Figure 13
Figure 13. Figure 13: Qualitative results of different network configurations on the three datasets. [PITH_FULL_IMAGE:figures/full_fig_p012_13.png] view at source ↗
Figure 14
Figure 14. Figure 14: Feature visualization results of different method on the TUE-CD dataset. [PITH_FULL_IMAGE:figures/full_fig_p013_14.png] view at source ↗
Figure 15
Figure 15. Figure 15: mF1-score of all methods for each epoch on the WHU-CD datasets [PITH_FULL_IMAGE:figures/full_fig_p014_15.png] view at source ↗
read the original abstract

As one of the most destructive natural disasters, earthquakes have struck many countries around the world in recent years, causing serious economic losses. Change detection (CD) can be applied to post-earthquake damage assessment as it can infer destroyed change regions from multi-temporal remote sensing images. Furthermore, the CD with short imaging interval will better satisfy the needs of the emergency rescues after earthquakes. However, the capability of current methods built on deep neural networks is limited because the dataset with short imaging interval is absent. To meet post-disaster immediate relief, we create a CD dataset, Turkey earthquake CD dataset (TUE-CD), for the evaluation of building damage in the short term after an earthquake. Because of the short acquisition interval of the post-event images, the imaging angle is different for different temporal images, which leads to some side-looking problems. To deal with these challenges, we present a multi-scale feature interaction network (MSI-Net) for efficient interaction between bi-temporal features, as well as mitigating the effect of side-looking problems. Specifically, the proposed MSI-Net consists of joint cross-attention (JCA) modules, multi-scale offset calibration (MOC) modules, and feature integration (FeI) modules. The JCA module unifies channel cross-attention and spatial joint attention for sufficient feature interaction. The MOC module further estimates the offsets to align the bi-temporal image with the multi-scale features. Finally, calibrated features and multi-scale features are fused by FeI modules for the prediction of changed areas. Experiments on the WHU-CD, CLCD, and the constructed TUE-CD dataset indicate that the proposed MSI-Net provides better results than considered state-of-the-art CD methods.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The paper introduces the TUE-CD dataset for building change detection using short-interval post-earthquake remote sensing images and proposes MSI-Net, which integrates Joint Cross-Attention (JCA) modules for feature interaction, Multi-Scale Offset Calibration (MOC) modules to estimate offsets addressing side-looking misalignment from differing imaging angles, and Feature Integration (FeI) modules for fusion. Experiments on WHU-CD, CLCD, and TUE-CD report superior performance over considered state-of-the-art change detection methods.

Significance. Creation of TUE-CD fills a documented gap in short-interval post-disaster datasets and could support emergency response applications if validated. The empirical gains on three datasets, including the new one, would be noteworthy if the MOC module's alignment benefit is confirmed; however, the absence of independent validation metrics limits the strength of the contribution.

major comments (2)
  1. [Experiments] Experiments section: the headline claim that MSI-Net outperforms SOTA on TUE-CD rests on the MOC module's ability to mitigate side-looking effects without introducing new registration errors, yet no ablation isolating MOC on TUE-CD, no quantitative alignment metrics (e.g., before/after offset error on building edges), and no qualitative feature-map comparisons are provided to verify that learned offsets are beneficial rather than harmful.
  2. [Method] Method section (MOC module description): the assumption that multi-scale offset estimation aligns bi-temporal features affected by short-interval angle differences is load-bearing for the TUE-CD results, but the paper provides no evidence or test that scale-specific estimation noise does not degrade performance on textured regions.
minor comments (2)
  1. [Abstract] Abstract and introduction: the phrase 'side-looking problems' is used without a concrete definition, example image pair, or citation to prior remote-sensing literature on off-nadir effects.
  2. [Dataset] Dataset section: details on how TUE-CD was constructed (e.g., exact acquisition dates, sensor, annotation protocol, train/val/test split sizes) are referenced but not fully specified, hindering reproducibility.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive comments, which highlight important aspects of validating the MOC module's contribution. We address each major comment below and commit to revisions that strengthen the empirical support without altering the core claims.

read point-by-point responses

  1. Referee: [Experiments] Experiments section: the headline claim that MSI-Net outperforms SOTA on TUE-CD rests on the MOC module's ability to mitigate side-looking effects without introducing new registration errors, yet no ablation isolating MOC on TUE-CD, no quantitative alignment metrics (e.g., before/after offset error on building edges), and no qualitative feature-map comparisons are provided to verify that learned offsets are beneficial rather than harmful.

    Authors: We agree that isolating the MOC module's contribution specifically on TUE-CD, along with quantitative alignment metrics and qualitative visualizations, would provide stronger evidence. In the revised version, we will add an ablation study removing MOC on TUE-CD, report before/after offset errors measured on building edges, and include qualitative feature-map comparisons demonstrating the alignment effect. revision: yes

  2. Referee: [Method] Method section (MOC module description): the assumption that multi-scale offset estimation aligns bi-temporal features affected by short-interval angle differences is load-bearing for the TUE-CD results, but the paper provides no evidence or test that scale-specific estimation noise does not degrade performance on textured regions.

    Authors: The multi-scale design of MOC is motivated by the need to handle offsets at different resolutions induced by angle differences, and the overall performance gains on TUE-CD are consistent with this. However, we acknowledge the lack of targeted tests for estimation noise on textured regions. We will incorporate additional analysis in the revision, such as offset visualizations on textured areas and performance comparisons to confirm no degradation occurs. revision: yes

Circularity Check

0 steps flagged

No significant circularity in empirical architecture and dataset proposal

full rationale

The paper proposes MSI-Net (with JCA, MOC, FeI modules) and constructs TUE-CD dataset, then reports empirical results on WHU-CD, CLCD and TUE-CD against SOTA baselines. No first-principles derivation, fitted parameter renamed as prediction, or self-citation load-bearing step is present; performance claims rest on standard training/evaluation rather than any reduction to inputs by construction. The work is self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 3 invented entities

The central claim depends on standard deep learning assumptions about feature alignment and interaction plus newly introduced modules whose effectiveness is validated only empirically on the presented datasets.

free parameters (1)
  • MSI-Net network parameters
    Weights and biases learned via training on the change detection datasets to achieve reported performance.
axioms (1)
  • domain assumption Bi-temporal remote sensing images with differing angles can be aligned via multi-scale offset estimation
    Invoked to justify the MOC module design and its role in mitigating side-looking problems.
invented entities (3)
  • Joint Cross-Attention (JCA) module no independent evidence
    purpose: Unify channel cross-attention and spatial joint attention for bi-temporal feature interaction
    Newly proposed component in the paper.
  • Multi-Scale Offset Calibration (MOC) module no independent evidence
    purpose: Estimate offsets to align bi-temporal images using multi-scale features
    Newly proposed to address side-looking issues.
  • Feature Integration (FeI) module no independent evidence
    purpose: Fuse calibrated multi-scale features for change area prediction
    Newly proposed for final output generation.

pith-pipeline@v0.9.1-grok · 5838 in / 1460 out tokens · 33660 ms · 2026-06-27T14:10:49.970588+00:00 · methodology

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

71 extracted references · 1 canonical work pages

  1. [1]

    Building-damage detection using pre- and post-seismic high-resolution satellite stereo imagery: A case study of the may 2008 wenchuan earthquake,

    X. Tong, Z. Hong, S. Liu, X. Zhang, H. Xie, Z. Li, S. Yang, W. Wang, and F. Bao, “Building-damage detection using pre- and post-seismic high-resolution satellite stereo imagery: A case study of the may 2008 wenchuan earthquake,”ISPRS J. Photogramm. Remote Sens., vol. 68, pp. 13–27, 2012. JOURNAL OF LATEX CLASS FILES, VOL. 14, NO. 8, AUGUST 2021 14 Fig. 15...

    2008

  2. [2]

    Automated machine learning driven stacked ensemble modeling for forest aboveground biomass prediction using multitemporal Sentinel-2 data,

    P. Naik, M. Dalponte, and L. Bruzzone, “Automated machine learning driven stacked ensemble modeling for forest aboveground biomass prediction using multitemporal Sentinel-2 data,”IEEE J. Sel. Topics Appl. Earth Observ. Remote Sens., vol. 16, pp. 3442–3454, 2023

    2023

  3. [3]

    A scalable high-performance unsupervised system for producing large-scale HR land cover maps: The Italian country case study,

    C. Paris, L. Gasparella, and L. Bruzzone, “A scalable high-performance unsupervised system for producing large-scale HR land cover maps: The Italian country case study,”IEEE J. Sel. Topics Appl. Earth Observ. Remote Sens., vol. 15, pp. 9146–9159, 2022

    2022

  4. [4]

    Cross-scale-guided fusion transformer for disaster assessment using satellite imagery,

    W. Xiao, J. Su, Y . Chen, and G. Cao, “Cross-scale-guided fusion transformer for disaster assessment using satellite imagery,”IEEE Trans. Geosci. Remote Sens., vol. 61, no. 4600512, pp. 1–12, 2023

    2023

  5. [5]

    Use of shadows for detection of earthquake-induced collapsed buildings in high-resolution satellite imagery,

    X. Tong, X. Lin, T. Feng, H. Xie, S. Liu, Z. Hong, and P. Chen, “Use of shadows for detection of earthquake-induced collapsed buildings in high-resolution satellite imagery,”ISPRS J. Photogramm. Remote Sens., vol. 79, pp. 53–67, 2013

    2013

  6. [6]

    Fully convolutional siamese networks for change detection,

    R. C. Daudt, B. Le Saux, and A. Boulch, “Fully convolutional siamese networks for change detection,”IEEE ICIP, pp. 4063–4067, 2018

    2018

  7. [7]

    Change detection on remote sensing images using dual-branch multilevel intertemporal network,

    Y . Feng, J. Jiang, H. Xu, and J. Zheng, “Change detection on remote sensing images using dual-branch multilevel intertemporal network,” IEEE Trans. Geosci. Remote Sens., vol. 61, no. 4401015, pp. 1–15, 2023

    2023

  8. [8]

    Remote sensing change detection via temporal feature interaction and guided refinement,

    Z. Li, C. Tang, L. Wang, and A. Y . Zomaya, “Remote sensing change detection via temporal feature interaction and guided refinement,”IEEE Trans. Geosci. Remote Sens., vol. 60, no. 5628711, pp. 1–11, 2022

    2022

  9. [9]

    Elgc-net: Efficient local–global context aggregation for remote sensing change detection,

    M. Noman, M. Fiaz, H. Cholakkal, S. Khan, and F. S. Khan, “Elgc-net: Efficient local–global context aggregation for remote sensing change detection,”IEEE Trans. Geosci. Remote Sens., vol. 62, no. 4701611, pp. 1–11, 2024

    2024

  10. [10]

    SASiamNet: Self- adaptive siamese network for change detection of remote sensing image,

    X. Long, W. Zhuang, M. Xia, K. Hu, and H. Lin, “SASiamNet: Self- adaptive siamese network for change detection of remote sensing image,” IEEE J. Sel. Topics Appl. Earth Observ. Remote Sens, vol. 17, pp. 1021– 1034, 2024

    2024

  11. [11]

    Changer: Feature interaction is what you need for change detection,

    S. Fang, K. Li, and Z. Li, “Changer: Feature interaction is what you need for change detection,”IEEE Trans. Geosci. Remote Sens., vol. 61, no. 5610111, pp. 1–11, 2023

    2023

  12. [12]

    Sdcdnet: A semi-dual change detection network framework with super- weak label for remote sensing image,

    J. Wang, F. Liu, H. Wang, X. Liu, L. Jiao, H. Yang, L. Li, and P. Chen, “Sdcdnet: A semi-dual change detection network framework with super- weak label for remote sensing image,”IEEE Trans. Geosci. Remote Sens., vol. 61, no. 5612714, pp. 1–14, 2023

    2023

  13. [13]

    Domain adaptive and interactive differential attention network for remote sensing image change detection,

    Y . Ji, W. Sun, Y . Wang, Z. Lv, G. Yang, Y . Zhan, and C. Li, “Domain adaptive and interactive differential attention network for remote sensing image change detection,”IEEE Trans. Geosci. Remote Sens., vol. 62, no. 5616316, pp. 1–16, 2024

    2024

  14. [14]

    Hyperboloid- embedded siamese network for change detection in remote sensing images,

    Q. Yang, S. Zhang, J. Li, Y . Sun, Q. Han, and Y . Sun, “Hyperboloid- embedded siamese network for change detection in remote sensing images,”IEEE J. Sel. Topics Appl. Earth Observ. Remote Sens, pp. 1–14, 2024

    2024

  15. [15]

    Content- guided spatial–spectral integration network for change detection in hr remote sensing images,

    Y . Liu, F. Zhang, S. Zhang, K. Zhang, J. Sun, and L. Bruzzone, “Content- guided spatial–spectral integration network for change detection in hr remote sensing images,”IEEE Trans. Geosci. Remote Sens., vol. 62, no. 5604816, pp. 1–16, 2024

    2024

  16. [16]

    Multiview hypergraph fusion network for change detection in high-resolution remote sensing images,

    X. Zhao, K. Zhang, F. Zhang, J. Sun, W. Wan, and H. Zhang, “Multiview hypergraph fusion network for change detection in high-resolution remote sensing images,”IEEE J. Sel. Topics Appl. Earth Observ. Remote Sens, vol. 17, pp. 4597–4610, 2024

    2024

  17. [17]

    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, no. 5607514, pp. 1–14, 2022

    2022

  18. [18]

    Remote sensing change detection with transformers trained from scratch,

    M. Noman, M. Fiaz, H. Cholakkal, S. Narayan, R. M. Anwer, S. Khan, and F. S. Khan, “Remote sensing change detection with transformers trained from scratch,”IEEE Trans. Geosci. Remote Sens., pp. 1–1, 2024

    2024

  19. [19]

    Relation changes matter: Cross-Temporal difference transformer for change detection in remote sensing images,

    K. Zhang, X. Zhao, F. Zhang, L. Ding, J. Sun, and L. Bruzzone, “Relation changes matter: Cross-Temporal difference transformer for change detection in remote sensing images,”IEEE Trans. Geosci. Remote Sens., vol. 61, no. 5611615, pp. 1–15, 2023

    2023

  20. [20]

    Gcformer: Global context-aware transformer for remote sensing image change detection,

    W. Yu, L. Zhuo, and J. Li, “Gcformer: Global context-aware transformer for remote sensing image change detection,”IEEE Trans. Geosci. Remote Sens., vol. 62, no. 4703212, pp. 1–12, 2024. JOURNAL OF LATEX CLASS FILES, VOL. 14, NO. 8, AUGUST 2021 15

    2024

  21. [21]

    Creating xbd: A dataset for assessing building damage from satellite imagery,

    R. Gupta, B. Goodman, N. Patel, R. Hosfelt, S. Sajeev, E. Heim, J. Doshi, K. Lucas, H. Choset, and M. Gaston, “Creating xbd: A dataset for assessing building damage from satellite imagery,” inIEEE CVPRW, 2019, pp. 10–17

    2019

  22. [22]

    Deep residual learning for image recognition,

    K. He, X. Zhang, S. Ren, and J. Sun, “Deep residual learning for image recognition,” inIEEE CVPR, 2016, pp. 770–778

    2016

  23. [23]

    Monitoring urban land cover changes at the urban fringe from spot hrv imagery in south-east england,

    N. Quarmby and J. Cushnie, “Monitoring urban land cover changes at the urban fringe from spot hrv imagery in south-east england,”Int. J. Remote Sens., vol. 10, no. 6, pp. 953–963, 1989

    1989

  24. [24]

    Procedures for change detection using landsat digital data,

    P. J. Howarth and G. M. Wickware, “Procedures for change detection using landsat digital data,”Int. J. Remote Sens., vol. 2, no. 3, pp. 277– 291, 1981

    1981

  25. [25]

    A theoretical framework for unsupervised change detection based on change vector analysis in the polar domain,

    F. Bovolo and L. Bruzzone, “A theoretical framework for unsupervised change detection based on change vector analysis in the polar domain,” IEEE Trans. Geosci. Remote. Sens., vol. 45, no. 1, pp. 218–236, 2006

    2006

  26. [26]

    A framework for automatic and unsupervised detection of multiple changes in multitemporal im- ages,

    F. Bovolo, S. Marchesi, and L. Bruzzone, “A framework for automatic and unsupervised detection of multiple changes in multitemporal im- ages,”IEEE Trans. Geosci. Remote. Sens., vol. 50, no. 6, pp. 2196–2212, 2012

    2012

  27. [27]

    Unsupervised change detection of multispectral images based on pca and low-rank prior,

    W. Zhang, J. Li, F. Zhang, J. Sun, and K. Zhang, “Unsupervised change detection of multispectral images based on pca and low-rank prior,” IEEE Geosci. Remote Sens. Lett., vol. 19, no. 5001605, pp. 1–5, 2022

    2022

  28. [28]

    Sedanet: A new siamese ensemble difference attention network for building change detection in remotely sensed images,

    Y . Yang, T. Chen, T. Lei, B. Du, A. K. Nandi, and A. Plaza, “Sedanet: A new siamese ensemble difference attention network for building change detection in remotely sensed images,”IEEE Trans. Geosci. Remote Sens., vol. 62, no. 4405316, pp. 1–16, 2024

    2024

  29. [29]

    Adjacent-level feature cross-fusion with 3-d cnn for remote sensing image change detection,

    Y . Ye, M. Wang, L. Zhou, G. Lei, J. Fan, and Y . Qin, “Adjacent-level feature cross-fusion with 3-d cnn for remote sensing image change detection,”IEEE Trans. Geosci. Remote Sens., vol. 61, no. 5618214, pp. 1–14, 2023

    2023

  30. [30]

    Deep depthwise separable convolutional network for change detection in optical aerial images,

    R. Liu, D. Jiang, L. Zhang, and Z. Zhang, “Deep depthwise separable convolutional network for change detection in optical aerial images,” IEEE J. Sel. Topics Appl. Earth Observ. Remote Sens, vol. 13, pp. 1109– 1118, 2020

    2020

  31. [31]

    Unsupervised cd in satellite image time series by contrastive learning and feature tracking,

    Y . Chen and L. Bruzzone, “Unsupervised cd in satellite image time series by contrastive learning and feature tracking,”IEEE Trans. Geosci. Remote Sens., vol. 62, no. 5401813, pp. 1–13, 2024

    2024

  32. [32]

    Bifa: Remote sensing image change detection with bitemporal feature alignment,

    H. Zhang, H. Chen, C. Zhou, K. Chen, C. Liu, Z. Zou, and Z. Shi, “Bifa: Remote sensing image change detection with bitemporal feature alignment,”IEEE Trans. Geosci. Remote Sens., vol. 62, no. 5614317, pp. 1–17, 2024

    2024

  33. [33]

    A self-supervised approach to pixel-level change detection in bi-temporal rs images,

    Y . Chen and L. Bruzzone, “A self-supervised approach to pixel-level change detection in bi-temporal rs images,”IEEE Trans. Geosci. Remote Sens., vol. 60, no. 4413911, pp. 1–11, 2022

    2022

  34. [34]

    S2dbft: Spectral-spatial dual-branch fusion transformer for hyperspectral image classification,

    Y . Zhang, Z. Wang, M. Huang, M. Li, J. Zhang, S. Wang, J. Zhang, and H. Zhang, “S2dbft: Spectral-spatial dual-branch fusion transformer for hyperspectral image classification,”IEEE Transactions on Geoscience and Remote Sensing, 2025

    2025

  35. [35]

    Implementation of motion estimation based on heterogeneous parallel computing system with opencl,

    J. Zhang, J.-F. Nezan, and J.-G. Cousin, “Implementation of motion estimation based on heterogeneous parallel computing system with opencl,” in2012 IEEE 14th International Conference on High Perfor- mance Computing and Communication & 2012 IEEE 9th International Conference on Embedded Software and Systems. IEEE, 2012, pp. 41– 45

    2012

  36. [36]

    3d octave and 2d vanilla mixed convolutional neural network for hyperspectral image classification with limited samples,

    Y . Feng, J. Zheng, M. Qin, C. Bai, and J. Zhang, “3d octave and 2d vanilla mixed convolutional neural network for hyperspectral image classification with limited samples,”Remote Sensing, vol. 13, no. 21, p. 4407, 2021

    2021

  37. [37]

    Learning vertex representations for bipartite networks,

    M. Gao, X. He, L. Chen, T. Liu, J. Zhang, and A. Zhou, “Learning vertex representations for bipartite networks,”IEEE transactions on knowledge and data engineering, vol. 34, no. 1, pp. 379–393, 2020

    2020

  38. [38]

    Multi-granularity episodic contrastive learning for few-shot learning,

    P. Zhu, Z. Zhu, Y . Wang, J. Zhang, and S. Zhao, “Multi-granularity episodic contrastive learning for few-shot learning,”Pattern Recognition, vol. 131, p. 108820, 2022

    2022

  39. [39]

    A novel ground-based cloud image segmentation method by using deep transfer learning,

    Z. Zhou, F. Zhang, H. Xiao, F. Wang, X. Hong, K. Wu, and J. Zhang, “A novel ground-based cloud image segmentation method by using deep transfer learning,”IEEE Geoscience and Remote Sensing Letters, vol. 19, pp. 1–5, 2021

    2021

  40. [40]

    Ensemble meteorological cloud classification meets internet of dependable and controllable things,

    J. Zhang, P. Liu, F. Zhang, H. Iwabuchi, A. A. d. H. e Ayres, V . H. C. De Albuquerqueet al., “Ensemble meteorological cloud classification meets internet of dependable and controllable things,”IEEE Internet of Things Journal, vol. 8, no. 5, pp. 3323–3330, 2020

    2020

  41. [41]

    Automated cca-mwf algorithm for unsupervised identification and removal of eog artifacts from eeg,

    M. Miao, W. Hu, B. Xu, J. Zhang, J. J. Rodrigues, and V . H. C. De Albu- querque, “Automated cca-mwf algorithm for unsupervised identification and removal of eog artifacts from eeg,”IEEE Journal of Biomedical and Health Informatics, vol. 26, no. 8, pp. 3607–3617, 2021

    2021

  42. [42]

    Supervised learning based discrete hashing for image retrieval,

    Q. Ma, C. Bai, J. Zhang, Z. Liu, and S. Chen, “Supervised learning based discrete hashing for image retrieval,”Pattern Recognition, vol. 92, pp. 156–164, 2019

    2019

  43. [43]

    Clothing sale forecasting by a composite gru–prophet model with an attention mechanism,

    Y . Li, Y . Yang, K. Zhu, and J. Zhang, “Clothing sale forecasting by a composite gru–prophet model with an attention mechanism,”IEEE Transactions on Industrial Informatics, vol. 17, no. 12, pp. 8335–8344, 2021

    2021

  44. [44]

    Distilled large language model-driven dynamic sparse expert activation mechanism,

    Q. Chen, Z. Zhang, Z. Zhang, K. Zhang, D. Li, W. Wang, J. Zhang, and C. Liu, “Distilled large language model-driven dynamic sparse expert activation mechanism,”Applied Soft Computing, p. 114037, 2025

    2025

  45. [45]

    Dual-path aggregation transformer network for super-resolution with images occlusions and variability,

    Q. Chen, L. Wang, Z. Zhang, X. Wang, W. Liu, B. Xia, H. Ding, J. Zhang, S. Xu, and X. Wang, “Dual-path aggregation transformer network for super-resolution with images occlusions and variability,” Engineering Applications of Artificial Intelligence, vol. 139, no. PartA, 2025

    2025

  46. [46]

    Kftd: Koopman- fourier time-differentiable network for continuous ocean spatiotemporal forecasting,

    Q. Chen, Z. Zhang, H. Liu, J. Zhang, and C. Bai, “Kftd: Koopman- fourier time-differentiable network for continuous ocean spatiotemporal forecasting,” inProceedings of the 32nd ACM SIGKDD Conference on Knowledge Discovery and Data Mining V . 1, 2026, pp. 94–103

    2026

  47. [47]

    A novel dataset and lightweight distil- lation baseline for highlight transparent object detection,

    Z. Zhang, G. Li, H. Zhang, Q. Chen, Q. Zhang, J. Wan, M. Xiong, C. Bai, D. Li, W. Zhanget al., “A novel dataset and lightweight distil- lation baseline for highlight transparent object detection,”International Journal of Computer Vision, vol. 134, no. 4, p. 157, 2026

    2026

  48. [48]

    Idd-net: Industrial defect detection method based on deep-learning,

    Z. Zhang, M. Zhou, H. Wan, M. Li, G. Li, and D. Han, “Idd-net: Industrial defect detection method based on deep-learning,”Engineering Applications of Artificial Intelligence, vol. 123, p. 106390, 2023

    2023

  49. [49]

    Zero-shot learning in industrial scenarios: New large- scale benchmark, challenges and baseline,

    Z. Zhang, Q. Chen, M. Xiong, S. Ding, Z. Su, X. Yao, Y . Sun, C. Bai, and J. Zhang, “Zero-shot learning in industrial scenarios: New large- scale benchmark, challenges and baseline,” inProceedings of the AAAI Conference on Artificial Intelligence, vol. 39, no. 10, 2025, pp. 10 357– 10 366

    2025

  50. [50]

    Representation learning based on co-evolutionary com- bined with probability distribution optimization for precise defect loca- tion,

    J. Zhang, Z. Zhang, Q. Chen, G. Li, W. Li, S. Ding, M. Xiong, W. Zhang, and S. Chen, “Representation learning based on co-evolutionary com- bined with probability distribution optimization for precise defect loca- tion,”IEEE Transactions on Neural Networks and Learning Systems, vol. 36, no. 7, pp. 11 989–12 003, 2024

    2024

  51. [51]

    Unification of closed-open industrial detection scenarios: New large-scale benchmarks, challenges and baselines,

    Z. Zhang, J. Zhang, Q. Chen, G. Li, D. Chen, S. Jing, H. Wang, D. Li, C. Liu, C. Baiet al., “Unification of closed-open industrial detection scenarios: New large-scale benchmarks, challenges and baselines,”IEEE Transactions on Pattern Analysis and Machine Intelligence, 2026

    2026

  52. [52]

    Cas-net: Comparison-based attention siamese network for change detection with an open high-resolution uav image dataset,

    Y . Zhai, W. Li, T. Xian, X. Jia, H. Zhang, Z. Tan, J. Zhou, J. Zeng, and C. L. Philip Chen, “Cas-net: Comparison-based attention siamese network for change detection with an open high-resolution uav image dataset,”IEEE Trans. Geosci. Remote Sens., vol. 62, no. 5619617, pp. 1–17, 2024

    2024

  53. [53]

    Beyond pixel-level annotation: Exploring self-supervised learning for change detection with image-level supervision,

    M. Zhao, X. Hu, L. Zhang, Q. Meng, Y . Chen, and L. Bruzzone, “Beyond pixel-level annotation: Exploring self-supervised learning for change detection with image-level supervision,”IEEE Trans. Geosci. Remote Sens., vol. 62, no. 5614916, pp. 1–16, 2024

    2024

  54. [54]

    A densely attentive refinement network for change detection based on very-high-resolution bitemporal remote sensing images,

    Z. Li, C. Yan, Y . Sun, and Q. Xin, “A densely attentive refinement network for change detection based on very-high-resolution bitemporal remote sensing images,”IEEE Trans. Geosci. Remote Sens., vol. 60, no. 4409818, pp. 1–18, 2022

    2022

  55. [55]

    Cit: Content-invariant translation with hybrid attention mechanism for unsupervised change detection,

    B. Fang, G. Chen, R. Kou, M. E. Paoletti, J. M. Haut, and A. Plaza, “Cit: Content-invariant translation with hybrid attention mechanism for unsupervised change detection,”ISPRS J. Photogramm. Remote Sens., vol. 204, pp. 321–339, 2023

    2023

  56. [56]

    Cascaded attention-induced difference representation learning for multispectral change detection,

    W. Zhang, Q. Zhang, H. Ning, and X. Lu, “Cascaded attention-induced difference representation learning for multispectral change detection,” Int. J. Appl. Earth Obs., vol. 121, p. 103366, 2023

    2023

  57. [57]

    Transunetcd: A hybrid transformer network for change detection in optical remote-sensing images,

    Q. Li, R. Zhong, X. Du, and Y . Du, “Transunetcd: A hybrid transformer network for change detection in optical remote-sensing images,”IEEE Trans. Geosci. Remote Sens., vol. 60, no. 5622519, pp. 1–19, 2022

    2022

  58. [58]

    A cnn-transformer network with multiscale context aggregation for fine-grained cropland change detection,

    M. Liu, Z. Chai, H. Deng, and R. Liu, “A cnn-transformer network with multiscale context aggregation for fine-grained cropland change detection,”IEEE J. Sel. Topics Appl. Earth Observ. Remote Sens, vol. 15, pp. 4297–4306, 2022

    2022

  59. [59]

    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 Trans. Geosci. Remote Sens., vol. 60, no. 5224713, pp. 1–13, 2022

    2022

  60. [60]

    Adaptformer: An adaptive hierarchical semantic ap- proach for change detection on remote sensing images,

    T. Huang, Y . Hong, Y . Pang, J. Liang, J. Hong, L. Huang, Y . Zhang, Y . Jia, and P. Savi, “Adaptformer: An adaptive hierarchical semantic ap- proach for change detection on remote sensing images,”IEEE Transac. Instru. Measur., vol. 73, no. 5502612, pp. 1–1, 2024

    2024

  61. [61]

    An unsupervised transformer-based multivariate alteration detection approach for change detection in vhr remote sensing images,

    Y . Lin, S. Liu, Y . Zheng, X. Tong, H. Xie, H. Zhu, K. Du, H. Zhao, and J. Zhang, “An unsupervised transformer-based multivariate alteration detection approach for change detection in vhr remote sensing images,” JOURNAL OF LATEX CLASS FILES, VOL. 14, NO. 8, AUGUST 2021 16 IEEE J. Sel. Topics Appl. Earth Observ. Remote Sens, vol. 17, pp. 1–10, 2024

    2021

  62. [62]

    Joint spatio-temporal modeling for semantic change detection in remote sens- ing images,

    L. Ding, J. Zhang, H. Guo, K. Zhang, B. Liu, and L. Bruzzone, “Joint spatio-temporal modeling for semantic change detection in remote sens- ing images,”IEEE Trans. Geosci. Remote Sens., vol. 62, no. 5610814, pp. 1–14, 2024

    2024

  63. [63]

    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 Recognit., vol. 132, p. 108960, 2022

    2022

  64. [64]

    Ddpm-cd: Remote sensing change detection using denoising diffusion probabilistic mod- els,

    W. G. C. Bandara, N. G. Nair, and V . M. Patel, “Ddpm-cd: Remote sensing change detection using denoising diffusion probabilistic mod- els,”arXiv preprint arXiv:2206.11892, 2022

    work page arXiv 2022

  65. [65]

    Leveraging diffusion modeling for remote sensing change detection in built-up urban areas,

    R. Wan, J. Zhang, Y . Huang, Y . Li, B. Hu, and B. Wang, “Leveraging diffusion modeling for remote sensing change detection in built-up urban areas,”IEEE Access, vol. 12, pp. 7028–7039, 2024

    2024

  66. [66]

    Deep learning for change detection in remote sensing: A review,

    T. Bai, L. Wang, D. Yinet al., “Deep learning for change detection in remote sensing: A review,”Geo spat. Inf. Sci., pp. 1–27, 2022

    2022

  67. [67]

    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

  68. [68]

    Building change detection for remote sensing images using a dual-task constrained deep siamese convolutional network model,

    Y . Liu, C. Pang, Z. Zhan, X. Zhang, and X. Yang, “Building change detection for remote sensing images using a dual-task constrained deep siamese convolutional network model,”IEEE Geosci. Remote Sens. Lett., vol. 18, no. 5, pp. 811–815, 2020

    2020

  69. [69]

    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

  70. [70]

    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. Nandi, “Ultralightweight spatial–spectral feature cooperation network for change detection in remote sensing images,”IEEE Trans. Geosci. Remote Sens., vol. 61, no. 4402114, pp. 1–14, 2023

    2023

  71. [71]

    Hanet: A hierarchical attention network for change detection with bitemporal very-high- resolution remote sensing images,

    C. Han, C. Wu, H. Guo, M. Hu, and H. Chen, “Hanet: A hierarchical attention network for change detection with bitemporal very-high- resolution remote sensing images,”IEEE J. Sel. Topics Appl. Earth Observ. Remote Sens, vol. 16, pp. 3867–3878, 2023

    2023