Home > Research > Publications & Outputs > Recurrent Semantic Change Detection in VHR Remo...

Electronic data

Links

Text available via DOI:

View graph of relations

Recurrent Semantic Change Detection in VHR Remote Sensing Images Using Visual Foundation Models

Research output: Contribution to Journal/MagazineJournal articlepeer-review

E-pub ahead of print

Standard

Recurrent Semantic Change Detection in VHR Remote Sensing Images Using Visual Foundation Models. / Zhang, J.; Ding, L.; Zhou, T. et al.
In: IEEE Transactions on Geoscience and Remote Sensing, Vol. 63, 5402314, 31.12.2025.

Research output: Contribution to Journal/MagazineJournal articlepeer-review

Harvard

Zhang, J, Ding, L, Zhou, T, Wang, J, Atkinson, PM & Bruzzone, L 2025, 'Recurrent Semantic Change Detection in VHR Remote Sensing Images Using Visual Foundation Models', IEEE Transactions on Geoscience and Remote Sensing, vol. 63, 5402314. https://doi.org/10.1109/TGRS.2025.3546808

APA

Zhang, J., Ding, L., Zhou, T., Wang, J., Atkinson, P. M., & Bruzzone, L. (2025). Recurrent Semantic Change Detection in VHR Remote Sensing Images Using Visual Foundation Models. IEEE Transactions on Geoscience and Remote Sensing, 63, Article 5402314. Advance online publication. https://doi.org/10.1109/TGRS.2025.3546808

Vancouver

Zhang J, Ding L, Zhou T, Wang J, Atkinson PM, Bruzzone L. Recurrent Semantic Change Detection in VHR Remote Sensing Images Using Visual Foundation Models. IEEE Transactions on Geoscience and Remote Sensing. 2025 Dec 31;63:5402314. Epub 2025 Mar 17. doi: 10.1109/TGRS.2025.3546808

Author

Zhang, J. ; Ding, L. ; Zhou, T. et al. / Recurrent Semantic Change Detection in VHR Remote Sensing Images Using Visual Foundation Models. In: IEEE Transactions on Geoscience and Remote Sensing. 2025 ; Vol. 63.

Bibtex

@article{1ff6b6a156104f01b1c518841fe30466,
title = "Recurrent Semantic Change Detection in VHR Remote Sensing Images Using Visual Foundation Models",
abstract = "Semantic change detection (SCD) involves the simultaneous extraction of changed regions and their corresponding semantic classifications (pre- and post-change) in remote sensing images (RSIs). Despite recent advancements in vision foundation models (VFMs), the fast-segment anything model has demonstrated insufficient performance in SCD. In this article, we propose a novel VFMs architecture for SCD, designated as VFM-ReSCD. This architecture integrates a side adapter (SA) into the VFM-ReSCD to fine-tune the fast segment anything model (FastSAM) network, enabling zero-shot transfer to novel image distributions and tasks. This enhancement facilitates the extraction of spatial features from very high-resolution (VHR) RSIs. Moreover, we introduce a recurrent neural network (RNN) to model semantic correlation and capture feature changes. We evaluated the proposed methodology on two benchmark datasets. Extensive experiments show that our method achieves state-of-the-art (SOTA) performances over existing approaches and outperforms other CNN-based methods on two RSI datasets.",
author = "J. Zhang and L. Ding and T. Zhou and J. Wang and P.M. Atkinson and L. Bruzzone",
year = "2025",
month = mar,
day = "17",
doi = "10.1109/TGRS.2025.3546808",
language = "English",
volume = "63",
journal = "IEEE Transactions on Geoscience and Remote Sensing",
issn = "0196-2892",
publisher = "IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC",

}

RIS

TY - JOUR

T1 - Recurrent Semantic Change Detection in VHR Remote Sensing Images Using Visual Foundation Models

AU - Zhang, J.

AU - Ding, L.

AU - Zhou, T.

AU - Wang, J.

AU - Atkinson, P.M.

AU - Bruzzone, L.

PY - 2025/3/17

Y1 - 2025/3/17

N2 - Semantic change detection (SCD) involves the simultaneous extraction of changed regions and their corresponding semantic classifications (pre- and post-change) in remote sensing images (RSIs). Despite recent advancements in vision foundation models (VFMs), the fast-segment anything model has demonstrated insufficient performance in SCD. In this article, we propose a novel VFMs architecture for SCD, designated as VFM-ReSCD. This architecture integrates a side adapter (SA) into the VFM-ReSCD to fine-tune the fast segment anything model (FastSAM) network, enabling zero-shot transfer to novel image distributions and tasks. This enhancement facilitates the extraction of spatial features from very high-resolution (VHR) RSIs. Moreover, we introduce a recurrent neural network (RNN) to model semantic correlation and capture feature changes. We evaluated the proposed methodology on two benchmark datasets. Extensive experiments show that our method achieves state-of-the-art (SOTA) performances over existing approaches and outperforms other CNN-based methods on two RSI datasets.

AB - Semantic change detection (SCD) involves the simultaneous extraction of changed regions and their corresponding semantic classifications (pre- and post-change) in remote sensing images (RSIs). Despite recent advancements in vision foundation models (VFMs), the fast-segment anything model has demonstrated insufficient performance in SCD. In this article, we propose a novel VFMs architecture for SCD, designated as VFM-ReSCD. This architecture integrates a side adapter (SA) into the VFM-ReSCD to fine-tune the fast segment anything model (FastSAM) network, enabling zero-shot transfer to novel image distributions and tasks. This enhancement facilitates the extraction of spatial features from very high-resolution (VHR) RSIs. Moreover, we introduce a recurrent neural network (RNN) to model semantic correlation and capture feature changes. We evaluated the proposed methodology on two benchmark datasets. Extensive experiments show that our method achieves state-of-the-art (SOTA) performances over existing approaches and outperforms other CNN-based methods on two RSI datasets.

U2 - 10.1109/TGRS.2025.3546808

DO - 10.1109/TGRS.2025.3546808

M3 - Journal article

VL - 63

JO - IEEE Transactions on Geoscience and Remote Sensing

JF - IEEE Transactions on Geoscience and Remote Sensing

SN - 0196-2892

M1 - 5402314

ER -