Deep learning-based object detection of offshore platforms on Sentinel-1 Imagery and the impact of synthetic training data

• URL: http://arxiv.org/abs/2511.04304v1
• Date: Thu, 06 Nov 2025 12:13:53 GMT
• Title: Deep learning-based object detection of offshore platforms on Sentinel-1 Imagery and the impact of synthetic training data
• Authors: Robin Spanier, Thorsten Hoeser, Claudia Kuenzer,
• Abstract summary: Development of robust models for offshore infrastructure detection relies on comprehensive, balanced datasets.<n>By training deep learning-based YOLOv10 object detection models with a combination of synthetic and real Sentinel-1 satellite imagery, this study investigates the use of synthetic data to enhance model performance.<n>In total, 3,529 offshore platforms were detected, including 411 in the North Sea, 1,519 in the Gulf of Mexico, and 1,593 in the Persian Gulf.
• Score: 0.3823356975862005
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: The recent and ongoing expansion of marine infrastructure, including offshore wind farms, oil and gas platforms, artificial islands, and aquaculture facilities, highlights the need for effective monitoring systems. The development of robust models for offshore infrastructure detection relies on comprehensive, balanced datasets, but falls short when samples are scarce, particularly for underrepresented object classes, shapes, and sizes. By training deep learning-based YOLOv10 object detection models with a combination of synthetic and real Sentinel-1 satellite imagery acquired in the fourth quarter of 2023 from four regions (Caspian Sea, South China Sea, Gulf of Guinea, and Coast of Brazil), this study investigates the use of synthetic training data to enhance model performance. We evaluated this approach by applying the model to detect offshore platforms in three unseen regions (Gulf of Mexico, North Sea, Persian Gulf) and thereby assess geographic transferability. This region-holdout evaluation demonstrated that the model generalises beyond the training areas. In total, 3,529 offshore platforms were detected, including 411 in the North Sea, 1,519 in the Gulf of Mexico, and 1,593 in the Persian Gulf. The model achieved an F1 score of 0.85, which improved to 0.90 upon incorporating synthetic data. We analysed how synthetic data enhances the representation of unbalanced classes and overall model performance, taking a first step toward globally transferable detection of offshore infrastructure. This study underscores the importance of balanced datasets and highlights synthetic data generation as an effective strategy to address common challenges in remote sensing, demonstrating the potential of deep learning for scalable, global offshore infrastructure monitoring.

Related papers

• OceanSAR-2: A Universal Feature Extractor for SAR Ocean Observation [55.978228064498865]
  We present OceanSAR-2, the second generation of our foundation model for SAR-based ocean observation.<n>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.
  arXiv  Detail & Related papers  (2026-01-12T10:20:43Z)
• Neural ocean forecasting from sparse satellite-derived observations: a case-study for SSH dynamics and altimetry data [25.95895236084694]
  We present an end-to-end deep learning framework for short-term forecasting of global sea surface dynamics based on sparse satellite altimetry data.<n>Our framework is developed within the OceanBench initiative, promoting standardized evaluation in ocean machine learning.
  arXiv  Detail & Related papers  (2025-12-15T11:28:03Z)
• OceanForecastBench: A Benchmark Dataset for Data-Driven Global Ocean Forecasting [10.30935627811107]
  Data-driven deep learning-based ocean forecast models have demonstrated significant potential in capturing complex ocean dynamics.<n>Despite these advancements, the absence of open-source benchmarks has led to inconsistent data usage and evaluation methods.<n>We propose OceanForecastBench, a benchmark offering three core contributions.
  arXiv  Detail & Related papers  (2025-11-24T03:57:43Z)
• Exploring the Underwater World Segmentation without Extra Training [55.291219073365546]
  We introduce textbfAquaOV255, the first large-scale and fine-grained underwater segmentation dataset.<n>We also present textbfEarth2Ocean, a training-free OV segmentation framework.
  arXiv  Detail & Related papers  (2025-11-11T07:22:56Z)
• Advancing Ocean State Estimation with efficient and scalable AI [22.24444646069193]
  We present an AI-driven Data Assimilation Framework for Ocean (ADAF-Ocean) that assimilates multi-source and multi-scale data.<n>ADAF-Ocean learns a continuous mapping from heterogeneous inputs to ocean states, preserving native data fidelity.
  arXiv  Detail & Related papers  (2025-11-08T15:24:23Z)
• A Sentinel-3 foundation model for ocean colour [8.571925606193703]
  We describe a new foundation model using the Prithvi-EO Vision Transformer architecture which has been pre-trained to reconstruct data from the Sentinel-3 Ocean and Land Colour Instrument (OLCI)<n>We evaluate the model by fine-tuning on two downstream marine earth observation tasks.<n>We conclude that this new generation of geospatial AI models has the potential to provide more robust, data-driven insights into ocean ecosystems and their role in global climate processes.
  arXiv  Detail & Related papers  (2025-09-25T15:00:38Z)
• Neptune-X: Active X-to-Maritime Generation for Universal Maritime Object Detection [54.1960918379255]
  Neptune-X is a data-centric generative-selection framework for maritime object detection.<n>X-to-Maritime is a multi-modality-conditioned generative model that synthesizes diverse and realistic maritime scenes.<n>Our approach sets a new benchmark in maritime scene synthesis, significantly improving detection accuracy.
  arXiv  Detail & Related papers  (2025-09-25T04:59:02Z)
• Towards Training-Free Underwater 3D Object Detection from Sonar Point Clouds: A Comparison of Traditional and Deep Learning Approaches [0.0]
  We develop and compare two paradigms for training-free detection of artificial structures in multibeam echo-sounder point clouds.<n>Our dual approach combines a physics-based sonar simulation pipeline that generates synthetic training data for state-of-the-art neural networks, with a robust model-based template matching system.<n>Our findings challenge conventional wisdom about data-hungry deep learning in underwater domains and establish the first large-scale benchmark for training-free underwater 3D detection.
  arXiv  Detail & Related papers  (2025-08-22T12:08:21Z)
• Generative Lagrangian data assimilation for ocean dynamics under extreme sparsity [0.0]
  Reconstructing ocean dynamics from observational data is fundamentally limited by the sparse, irregular, and Lagrangian nature of spatial sampling.<n>We leverage a deep learning framework that combines neural operators with denoising diffusion probabilistic models (DDPMs) to reconstruct high-resolution ocean states.<n>We validate our method on benchmark systems, synthetic float observations, and real satellite data, demonstrating robust performance under severe spatial sampling limitations.
  arXiv  Detail & Related papers  (2025-07-09T01:56:25Z)
• ReconMOST: Multi-Layer Sea Temperature Reconstruction with Observations-Guided Diffusion [48.540756751934836]
  ReconMOST is a data-driven guided diffusion model framework for multi-layer sea temperature reconstruction.<n>Our method extends ML-based SST reconstruction to a global, multi-layer setting, handling over 92.5% missing data.
  arXiv  Detail & Related papers  (2025-06-12T06:27:22Z)
• Efficient Self-Supervised Learning for Earth Observation via Dynamic Dataset Curation [67.23953699167274]
  Self-supervised learning (SSL) has enabled the development of vision foundation models for Earth Observation (EO)<n>In EO, this challenge is amplified by the redundancy and heavy-tailed distributions common in satellite imagery.<n>We propose a dynamic dataset pruning strategy designed to improve SSL pre-training by maximizing dataset diversity and balance.
  arXiv  Detail & Related papers  (2025-04-09T15:13:26Z)
• Tackling water table depth modeling via machine learning: From proxy observations to verifiability [0.40964539027092917]
  Spatial patterns of water table depth (WTD) play a crucial role in shaping ecological resilience, hydrological connectivity, and human-centric systems.<n>We construct three fine-resolution (500 m) ML simulations of WTD using the XGBoost algorithm and more than 20 million real and proxy observations of WTD.<n>Our models can more accurately predict unseen real and proxy observations of WTD compared to two available PB simulations of WTD.
  arXiv  Detail & Related papers  (2024-04-30T09:05:17Z)
• Learning-based estimation of in-situ wind speed from underwater acoustics [58.293528982012255]
  We introduce a deep learning approach for the retrieval of wind speed time series from underwater acoustics. Our approach bridges data assimilation and learning-based frameworks to benefit both from prior physical knowledge and computational efficiency.
  arXiv  Detail & Related papers  (2022-08-18T15:27:40Z)
• Real-Time target detection in maritime scenarios based on YOLOv3 model [65.35132992156942]
  A novel ships dataset is proposed consisting of more than 56k images of marine vessels collected by means of web-scraping. A YOLOv3 single-stage detector based on Keras API is built on top of this dataset.
  arXiv  Detail & Related papers  (2020-02-10T15:25:19Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.