OceanSAR-1-Wind is a linear probing head for wind speed prediction built on top of the OceanSAR-1 foundation model. It leverages the powerful features extracted by OceanSAR-1 to accurately predict wind speed from Synthetic Aperture Radar (SAR) imagery.
This model is designed for wind speed prediction from SAR imagery, particularly over ocean surfaces. It can be used for:
The model achieves state-of-the-art linear probing performances on wind speed prediction, with performance varying by backbone architecture:
1import torch
2from transformers import AutoModelForImageClassification
3
4# Load the foundation model and wind prediction head
5oceansar = AutoModelForImageClassification.from_pretrained("galeio-research/OceanSAR-1-wind")
6
7# Prepare your SAR image (should be single-channel VV polarization)
8dummy_image = torch.randn(1, 1, 256, 256) # (C, H, W)
9
10# Extract features and predict wind speed
11with torch.no_grad():
12 wind_speed = oceansar(dummy_image).logits # Output in m/s
Wind speed prediction performance is evaluated using Root Mean Square Error (RMSE), achieving:
1@article{kerdreux2025efficientselfsupervisedlearningearth,
2 title={Efficient Self-Supervised Learning for Earth Observation via Dynamic Dataset Curation},
3 author={Kerdreux, Thomas and Tuel, Alexandre and Febvre, Quentin and Mouche, Alexis and Chapron, Bertrand},
4 journal={arXiv preprint arXiv:2504.06962},
5 year={2025},
6 eprint={2504.06962},
7 archivePrefix={arXiv},
8 primaryClass={cs.CV},
9 url={https://arxiv.org/abs/2504.06962},
10}
This work was granted access to the HPC resources of IDRIS and TGCC under the allocation 2025-[A0171015666] made by GENCI.
