mlsanityData to 7 Oct 2026

SSL4EO-S12: A Large-Scale Multi-Modal, Multi-Temporal Dataset for Self-Supervised Learning in Earth Observation

arXiv 2211.07044 ↗First posted 13 Nov 2022cs.CV cs.AIcited by 82

Not on the map. Why?

Abstract (arXiv)

Self-supervised pre-training bears potential to generate expressive representations without human annotation. Most pre-training in Earth observation (EO) are based on ImageNet or medium-size, labeled remote sensing (RS) datasets. We share an unlabeled RS dataset SSL4EO-S12 (Self-Supervised Learning for Earth Observation - Sentinel-1/2) to assemble a large-scale, global, multimodal, and multi-seasonal corpus of satellite imagery from the ESA Sentinel-1 & -2 satellite missions. For EO applications we demonstrate SSL4EO-S12 to succeed in self-supervised pre-training for a set of methods: MoCo-v2, DINO, MAE, and data2vec. Resulting models yield downstream performance close to, or surpassing accuracy measures of supervised learning. In addition, pre-training on SSL4EO-S12 excels compared to existing datasets. We make openly available the dataset, related source code, and pre-trained models at https://github.com/zhu-xlab/SSL4EO-S12.

Key facts. Cited by 82 papers in the map (data through 7 Oct 2026). Builds on 36 of its 54 references that are in the map. First posted on arXiv 13 Nov 2022.

Builds on

36 of 54 references.

References · 36

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 29,915
  3. 2015U-Net: Convolutional Networks for Biomedical Image Segmentationcited by 19,604
  4. 2020A Simple Framework for Contrastive Learning of Visual Representationscited by 11,979
  5. 2018Representation Learning with Contrastive Predictive Codingcited by 7,662
  6. 2019Momentum Contrast for Unsupervised Visual Representation Learningcited by 7,482
  7. 2021Masked Autoencoders Are Scalable Vision Learnerscited by 7,339
  8. 2021Emerging Properties in Self-Supervised Vision Transformerscited by 6,893
  9. 2020Bootstrap your own latent: A new approach to self-supervised Learningcited by 4,712
  10. 2018Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentationcited by 3,210
  11. 2020Unsupervised Learning of Visual Features by Contrasting Cluster Assignmentscited by 2,845
  12. 2016Pyramid Scene Parsing Networkcited by 2,472
24 more
  1. 2020Improved Baselines with Momentum Contrastive Learningcited by 2,282
  2. 2021Barlow Twins: Self-Supervised Learning via Redundancy Reductioncited by 1,818
  3. 2021An Empirical Study of Training Self-Supervised Vision Transformerscited by 1,610
  4. 2017EuroSAT: A Novel Dataset and Deep Learning Benchmark for Land Use and Land Cover Classificationcited by 1,488
  5. 2018Datasheets for Datasetscited by 1,371
  6. 2021iBOT: Image BERT Pre-Training with Online Tokenizercited by 891
  7. 2022data2vec: A General Framework for Self-supervised Learning in Speech, Vision and Languagecited by 718
  8. 2020Self-supervised Learning: Generative or Contrastivecited by 693
  9. 2019BigEarthNet: A Large-Scale Benchmark Archive For Remote Sensing Image Understandingcited by 231
  10. 2021Seasonal Contrast: Unsupervised Pre-Training from Uncurated Remote Sensing Datacited by 214
  11. 2016What makes ImageNet good for transfer learning?cited by 192
  12. 2020Geography-Aware Self-Supervised Learningcited by 190
  13. 2022Self-supervised Learning in Remote Sensing: A Reviewcited by 137
  14. 2018Urban Change Detection for Multispectral Earth Observation Using Convolutional Neural Networkscited by 108
  15. 2019SEN12MS -- A Curated Dataset of Georeferenced Multi-Spectral Sentinel-1/2 Imagery for Deep Learning and Data Fusioncited by 95
  16. 2021When Does Contrastive Visual Representation Learning Work?cited by 92
  17. 2021Self-Supervised Learning of Remote Sensing Scene Representations Using Contrastive Multiview Codingcited by 58
  18. 2022Self-supervised Vision Transformers for Joint SAR-optical Representation Learningcited by 40
  19. 2022EarthNets: Empowering AI in Earth Observationcited by 40
  20. 2019So2Sat LCZ42: A Benchmark Dataset for Global Local Climate Zones Classificationcited by 24
  21. 2021Self-supervised Audiovisual Representation Learning for Remote Sensing Datacited by 24
  22. 2022Extreme Masking for Learning Instance and Distributed Visual Representationscited by 21
  23. 2021Self-Supervised Multisensor Change Detectioncited by 19
  24. 2020Training general representations for remote sensing using in-domain knowledgecited by 12

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2022: 3; 2023: 17; 2024: 19; 2025: 21; 2026: 22.

82 papers.

Sort:

By area:

  1. 2023SkySense: A Multi-Modal Remote Sensing Foundation Model Towards Universal Interpretation for Earth Observation Imagerycited by 202
  2. 2023CROMA: Remote Sensing Representations with Contrastive Radar-Optical Masked Autoencoderscited by 167
  3. 2022SatlasPretrain: A Large-Scale Dataset for Remote Sensing Image Understandingcited by 164
  4. 2023SatCLIP: Global, General-Purpose Location Embeddings with Satellite Imagerycited by 148
  5. 2023GEO-Bench: Toward Foundation Models for Earth Monitoringcited by 122
  6. 2024Neural Plasticity-Inspired Multimodal Foundation Model for Earth Observationcited by 97
  7. 2024MMEarth: Exploring Multi-Modal Pretext Tasks For Geospatial Representation Learningcited by 76
  8. 2024AnySat: One Earth Observation Model for Many Resolutions, Scales, and Modalitiescited by 73
  9. 2024OmniSat: Self-Supervised Modality Fusion for Earth Observationcited by 63
  10. 2023A Billion-scale Foundation Model for Remote Sensing Imagescited by 62
  11. 2025Towards a Unified Copernicus Foundation Model for Earth Visioncited by 52
  12. 2023SSL4EO-L: Datasets and Foundation Models for Landsat Imagerycited by 49
38 more
  1. 2023Feature Guided Masked Autoencoder for Self-supervised Learning in Remote Sensingcited by 48
  2. 2024SpectralEarth: Training Hyperspectral Foundation Models at Scalecited by 42
  3. 2025Galileo: Learning Global & Local Features of Many Remote Sensing Modalitiescited by 41
  4. 2023Decoupling Common and Unique Representations for Multimodal Self-supervised Learningcited by 36
  5. 2025TESSERA: Temporal Embeddings of Surface Spectra for Earth Representation and Analysiscited by 35
  6. 2023Charting New Territories: Exploring the Geographic and Geospatial Capabilities of Multimodal LLMscited by 34
  7. 2024One for All: Toward Unified Foundation Models for Earth Visioncited by 34
  8. 2025Panopticon: Advancing Any-Sensor Foundation Models for Earth Observationcited by 34
  9. 2024On the Foundations of Earth and Climate Foundation Modelscited by 31
  10. 2023Revisiting pre-trained remote sensing model benchmarks: resizing and normalization matterscited by 27
  11. 2025SkySense V2: A Unified Foundation Model for Multi-modal Remote Sensingcited by 26
  12. 2024PhilEO Bench: Evaluating Geo-Spatial Foundation Modelscited by 25
  13. 2022Self-supervised remote sensing feature learning: Learning Paradigms, Challenges, and Future Workscited by 24
  14. 2023Better, Not Just More: Data-Centric Machine Learning for Earth Observationcited by 19
  15. 2024Mind the Modality Gap: Towards a Remote Sensing Vision-Language Model via Cross-modal Alignmentcited by 15
  16. 2024A2-MAE: A spatial-temporal-spectral unified remote sensing pre-training method based on anchor-aware masked autoencodercited by 10
  17. 2025Fine-tune Smarter, Not Harder: Parameter-Efficient Fine-Tuning for Geospatial Foundation Modelscited by 9
  18. 2026No One Knows the State of the Art in Geospatial Foundation Modelscited by 9
  19. 2023Continuous Cross-resolution Remote Sensing Image Change Detectioncited by 8
  20. 2024QuakeSet: A Dataset and Low-Resource Models to Monitor Earthquakes through Sentinel-1cited by 8
  21. 2024Evaluating and Benchmarking Foundation Models for Earth Observation and Geospatial AIcited by 7
  22. 2024Fus-MAE: A cross-attention-based data fusion approach for Masked Autoencoders in remote sensingcited by 6
  23. 2024Can Distillation Mitigate Backdoor Attacks in Pre-trained Encoders?cited by 6
  24. 2023Ben-ge: Extending BigEarthNet with Geographical and Environmental Datacited by 5
  25. 2025SSL4Eco: A Global Seasonal Dataset for Geospatial Foundation Models in Ecologycited by 5
  26. 2023In-Domain Self-Supervised Learning Improves Remote Sensing Image Scene Classificationcited by 4
  27. 2025A Genealogy of Foundation Models in Remote Sensingcited by 4
  28. 2025GeoCrossBench: Cross-Band Generalization for Remote Sensingcited by 4
  29. 2026Earth Embeddings as Products: Taxonomy, Ecosystem, and Standardized Accesscited by 4
  30. 2023Robust Burned Area Delineation through Multitask Learningcited by 3
  31. 2023Zooming Out on Zooming In: Advancing Super-Resolution for Remote Sensingcited by 3
  32. 2025Lightweight Metadata-Aware Mixture-of-Experts Masked Autoencoder for Earth Observationcited by 3
  33. 2023Self-Supervised Learning for Improved Synthetic Aperture Sonar Target Recognitioncited by 2
  34. 2026BigEarthNet.txt: A Large-Scale Multi-Sensor Image-Text Dataset and Benchmark for Earth Observationcited by 2
  35. 2024Estimating Physical Information Consistency of Channel Data Augmentation for Remote Sensing Imagescited by 1
  36. 2024Neural Embedding Compression For Efficient Multi-Task Earth Observation Modellingcited by 1
  37. 2026NeighborMAE: Exploiting Spatial Dependencies between Neighboring Earth Observation Images in Masked Autoencoders Pretrainingcited by 1
  38. 2026How to Embed Matters: Evaluation of EO Embedding Design Choicescited by 1

50 most cited of 82 All 82 (JSON)