Metraipolis › archive › dc_d766bbd3ba43

SAR2Agri: Learning SAR Intensity Representations for Agricultural Monitoring — arXiv:2608.11142

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.11142
License (as archived)
CC0-1.0
Archived
2026-10-19T06:30:40.706Z
SHA-256 of the archived text
1bcddfbe59cc6fa34d0d91d6d1e35b6a6ef9e0e26e1c95bfbb865a7842bf993d
Stable URL of this page
https://metraipolis.com/documento/dc_d766bbd3ba43

How to cite

SAR2Agri: Learning SAR Intensity Representations for Agricultural Monitoring — arXiv:2608.11142 — arXiv (en), https://arxiv.org/abs/2608.11142 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_d766bbd3ba43 (Metraipolis archive dc_d766bbd3ba43, sha256 1bcddfbe59cc6fa34d0d91d6d1e35b6a6ef9e0e26e1c95bfbb865a7842bf993d)
BibTeX
@misc{dcd766bbd3ba43,
  title = {SAR2Agri: Learning SAR Intensity Representations for Agricultural Monitoring — arXiv:2608.11142},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.11142},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_d766bbd3ba43; sha256 1bcddfbe59cc6fa34d0d91d6d1e35b6a6ef9e0e26e1c95bfbb865a7842bf993d},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.11142. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Agricultural monitoring faces unique challenges, arising from the landscape's complex temporal, phenological, and climate dynamics, yet monitoring them is critical for ensuring food security. Synthetic Aperture Radar (SAR) satellites offer all-weather day-night imaging capability supporting key monitoring tasks including crop type mapping, yield prediction and phenological event detection. Existing multimodal remote sensing foundation models including TerraMind and CopernicusFM learn SAR representations by grounding them in optical imagery using joint encoding and contrastive learning techniques, while SAR-specific foundation models such as SAR-JEPA, SARMAE, and SAR-W-MixMAE primarily focus on target detection, flood mapping, and land cover classification applications. Recent work has introduced phenology inspired temporal pretext tasks with optical imagery which has shown strong performance on agricultural downstream tasks. In this work, we propose the first self-supervised learning pipeline focused on using only SAR intensity imagery for agricultural applications. We improve the temporal pretext tasks through masking and curriculum learning to enhance the pretraining pipeline's ability to capture phenological features from SAR. On the SICKLE benchmark, our final model achieves 84.9% IoU on crop type mapping, outperforming optical baselines (by 15.3 pt) and existing SAR baselines (by 2.2 pt), demonstrating the effectiveness of our proposed pipeline for pretraining SAR intensity encoders for agricultural monitoring.

Tema degli esploratori: statistica e scienza dei dati (retro-catalogo arXiv: annunci del 2026-08-12 in cs.CV).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV).
Pagina del lavoro: https://arxiv.org/abs/2608.11142 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-19T06:25:27.321Z dalla pagina https://arxiv.org/catchup/cs.CV/2026-08-12?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).