OceanSAR-2: A Universal Feature Extractor for SAR Ocean Observation

2026/01/12 by Alexandre Tuel, Thomas Kerdreux, Quentin Febvre +6 · 1 voice
Computer Science · #cs.AI #cs.CV #cs.LG

paper · pdf · doi:10.48550/arxiv.2601.07392

Abstract

We present OceanSAR-2, the second generation of our foundation model for SAR-based ocean observation. Building on our earlier release, which pioneered self-supervised learning on Sentinel-1 Wave Mode data, OceanSAR-2 relies on improved SSL training and dynamic data curation strategies, which enhances performance while reducing training cost. OceanSAR-2 demonstrates strong transfer performance across downstream tasks, including geophysical pattern classification, ocean surface wind vector and significant wave height estimation, and iceberg detection. We release standardized benchmark datasets, providing a foundation for systematic evaluation and advancement of SAR models for ocean applications.

Citations

• TerraMind: Large-Scale Generative Multimodality for Earth Observation
• TerraMesh: A Planetary Mosaic of Multimodal Earth Observation Data
• Efficient Self-Supervised Learning for Earth Observation via Dynamic Dataset Curation
• Towards a Unified Copernicus Foundation Model for Earth Vision
• Prithvi-EO-2.0: A Versatile Multi-Temporal Foundation Model for Earth Observation Applications
• DINOv2: Learning Robust Visual Features without Supervision
• SSL4EO-S12: A Large-Scale Multi-Modal, Multi-Temporal Dataset for Self-Supervised Learning in Earth Observation
• SatMAE: Pre-training Transformers for Temporal and Multi-Spectral Satellite Imagery
• iBOT: Image BERT Pre-Training with Online Tokenizer
• Emerging Properties in Self-Supervised Vision Transformers
• Deformable DETR: Deformable Transformers for End-to-End Object Detection
• End-to-End Object Detection with Transformers

Discussions

• "OceanSAR-2: A Universal Feature Extractor for SAR Ocean Observation" arxiv.org/abs/2601.07392 [bsky, 0 points, 0 comments]

Related

• Efficient Self-Supervised Learning for Earth Observation via Dynamic Dataset Curation
• xView3-SAR: Detecting Dark Fishing Activity Using Synthetic Aperture Radar Imagery
• WV-Net: A foundation model for SAR WV-mode satellite imagery trained using contrastive self-supervised learning on 10 million images
• Guided deep learning by subaperture decomposition: ocean patterns from SAR imagery
• Technical report on target classification in SAR track