ORIGINAL RESEARCH article
Front. Remote Sens.
Sec. Agro-Environmental Remote Sensing
SAU-MTF: Siamese Attention U-Net with Multimodal Temporal Fusion for Accurate Deforestation Detection
Provisionally accepted
- Vel Tech Rangarajan Dr.Sagunthala R&D Institute of Science and Technology, Chennai, India
Select one of your emails
You have multiple emails registered with Frontiers:
Notify me on publication
Please enter your email address:
If you already have an account, please login
You don't have a Frontiers account ? You can register here
Monitoring deforestation through satellite imagery plays a crucial role in achieving environmental sustainability and combating climate change. However, conventional models face limitations, particularly under cloud-covered conditions and when tracking changes over time. This paper introduces a novel approach named Siamese Attention U-Net with Multimodal Temporal Fusion (SAU-MTF), which leverages both optical (Sentinel-2) and SAR (Sentinel-1) data across a tri-temporal framework to enhance deforestation detection accuracy. The architecture incorporates EfficientNet-based encoders, attention gates for salient feature extraction, and temporal context blocks that model changes over time. A multimodal fusion strategy allows complementary use of SAR's cloud-penetrating capabilities and optical data's visual richness. Evaluated on large-scale deforestation datasets, the proposed SAU-MTF model achieves an accuracy of 94.7% and 0.93 IoU, outperforming several state-of-the-art models. The results demonstrate the effectiveness of combining temporal, spectral, and spatial information for robust forest monitoring, particularly in regions affected by seasonal variations and cloud obstructions.
Keywords: deep learning, Deforestation detection, Environmental Monitoring, multimodal fusion, remote sensing, SAR and Optical Imagery, Sentinel-1, Sentinel-2
Received: 08 Sep 2025; Accepted: 16 Feb 2026.
Copyright: © 2026 Priya. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
* Correspondence: Poovayar Priya
Disclaimer: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.