mlsanityData to 7 Oct 2026

SatMAE: Pre-training Transformers for Temporal and Multi-Spectral Satellite Imagery

arXiv 2207.08051 ↗First posted 17 Jul 2022cs.CV cs.AIcited by 402

Not on the map. Why?

Abstract (arXiv)

Unsupervised pre-training methods for large vision models have shown to enhance performance on downstream supervised tasks. Developing similar techniques for satellite imagery presents significant opportunities as unlabelled data is plentiful and the inherent temporal and multi-spectral structure provides avenues to further improve existing pre-training strategies. In this paper, we present SatMAE, a pre-training framework for temporal or multi-spectral satellite imagery based on Masked Autoencoder (MAE). To leverage temporal information, we include a temporal embedding along with independently masking image patches across time. In addition, we demonstrate that encoding multi-spectral data as groups of bands with distinct spectral positional encodings is beneficial. Our approach yields strong improvements over previous state-of-the-art techniques, both in terms of supervised learning performance on benchmark datasets (up to uparrow 7%), and transfer learning performance on downstream remote sensing tasks, including land cover classification (up to uparrow 14%) and semantic segmentation. Code and data are available on the project website: https://sustainlab-group.github.io/SatMAE/

Key facts. Cited by 402 papers in the map (data through 7 Oct 2026). Builds on 49 of its 71 references that are in the map. First posted on arXiv 17 Jul 2022.

Builds on

49 of 71 references.

References · 49

  1. 2017Attention Is All You Needcited by 59,932
  2. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  3. 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 29,915
  4. 2020A Simple Framework for Contrastive Learning of Visual Representationscited by 11,979
  5. 2019Momentum Contrast for Unsupervised Visual Representation Learningcited by 7,482
  6. 2021Masked Autoencoders Are Scalable Vision Learnerscited by 7,339
  7. 2020Bootstrap your own latent: A new approach to self-supervised Learningcited by 4,712
  8. 2020Training data-efficient image transformers & distillation through attentioncited by 4,560
  9. 2020Exploring Simple Siamese Representation Learningcited by 2,853
  10. 2020Unsupervised Learning of Visual Features by Contrasting Cluster Assignmentscited by 2,845
  11. 2021ViViT: A Video Vision Transformercited by 1,610
  12. 2021An Empirical Study of Training Self-Supervised Vision Transformerscited by 1,610
37 more
  1. 2017EuroSAT: A Novel Dataset and Deep Learning Benchmark for Land Use and Land Cover Classificationcited by 1,488
  2. 2015Convolutional LSTM Network: A Machine Learning Approach for Precipitation Nowcastingcited by 1,428
  3. 2022VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Trainingcited by 1,305
  4. 2019Contrastive Multiview Codingcited by 1,297
  5. 2021Video Swin Transformercited by 1,090
  6. 2019Self-Supervised Learning of Pretext-Invariant Representationscited by 806
  7. 2019Self-supervised Visual Feature Learning with Deep Neural Networks: A Surveycited by 656
  8. 2020A Survey on Contrastive Self-supervised Learningcited by 553
  9. 2019Quantifying the Carbon Emissions of Machine Learningcited by 516
  10. 2017Functional Map of the Worldcited by 338
  11. 2020Self-supervised Equivariant Attention Mechanism for Weakly Supervised Semantic Segmentationcited by 299
  12. 2022MultiMAE: Multi-modal Multi-task Masked Autoencoderscited by 287
  13. 2019Revisiting Self-Supervised Visual Representation Learningcited by 250
  14. 2019BigEarthNet: A Large-Scale Benchmark Archive For Remote Sensing Image Understandingcited by 231
  15. 2021Seasonal Contrast: Unsupervised Pre-Training from Uncurated Remote Sensing Datacited by 214
  16. 2020Geography-Aware Self-Supervised Learningcited by 190
  17. 2018SpaceNet: A Remote Sensing Dataset and Challenge Seriescited by 178
  18. 2018xView: Objects in Context in Overhead Imagerycited by 145
  19. 2021Panoptic Segmentation of Satellite Image Time Series with Convolutional Temporal Attention Networkscited by 141
  20. 2021D2C: Diffusion-Denoising Models for Few-shot Conditional Generationcited by 96
  21. 2018Temporal Convolutional Neural Network for the Classification of Satellite Image Time Seriescited by 76
  22. 2019Self-attention for raw optical Satellite Time Series Classificationcited by 75
  23. 2020Using satellite imagery to understand and promote sustainable developmentcited by 71
  24. 2020CompRess: Self-Supervised Learning by Compressing Representationscited by 61
  25. 2021Self-Supervised Learning of Remote Sensing Scene Representations Using Contrastive Multiview Codingcited by 58
  26. 2021SustainBench: Benchmarks for Monitoring the Sustainable Development Goals with Machine Learningcited by 53
  27. 2021Lossy Compression for Lossless Predictioncited by 51
  28. 2019In-domain representation learning for remote sensingcited by 41
  29. 2021Geographical Knowledge-driven Representation Learning for Remote Sensing Imagescited by 37
  30. 2020Generating Interpretable Poverty Maps using Object Detection in Satellite Imagescited by 27
  31. 2021Spatial-Temporal Super-Resolution of Satellite Imagery via Conditional Pixel Synthesiscited by 20
  32. 2021Representation Learning for Remote Sensing: An Unsupervised Sensor Fusion Approachcited by 14
  33. 2021IS-COUNT: Large-scale Object Counting from Satellite Images with Covariate-based Importance Samplingcited by 14
  34. 2018Learning to Interpret Satellite Images Using Wikipediacited by 12
  35. 2018Convolutional LSTMs for Cloud-Robust Segmentation of Remote Sensing Imagerycited by 11
  36. 2019Predicting Economic Development using Geolocated Wikipedia Articlescited by 11
  37. 2022Tracking Urbanization in Developing Regions with Remote Sensing Spatial-Temporal Super-Resolutioncited by 2

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2022: 2; 2023: 43; 2024: 82; 2025: 128; 2026: 147.

402 papers.

Sort:

By area:

  1. 2023RemoteCLIP: A Vision Language Foundation Model for Remote Sensingcited by 357
  2. 2023ClimaX: A foundation model for weather and climatecited by 325
  3. 2022Scale-MAE: A Scale-Aware Masked Autoencoder for Multiscale Geospatial Representation Learningcited by 302
  4. 2023SpectralGPT: Spectral Remote Sensing Foundation Modelcited by 250
  5. 2023SkySense: A Multi-Modal Remote Sensing Foundation Model Towards Universal Interpretation for Earth Observation Imagerycited by 202
  6. 2024EarthGPT: A Universal Multi-modal Large Language Model for Multi-sensor Image Comprehension in Remote Sensing Domaincited by 181
  7. 2023RSPrompter: Learning to Prompt for Remote Sensing Instance Segmentation based on Visual Foundation Modelcited by 180
  8. 2023CROMA: Remote Sensing Representations with Contrastive Radar-Optical Masked Autoencoderscited by 167
  9. 2023Foundation Models for Generalist Geospatial Artificial Intelligencecited by 143
  10. 2024Prithvi-EO-2.0: A Versatile Multi-Temporal Foundation Model for Earth Observation Applicationscited by 139
  11. 2023DiffusionSat: A Generative Foundation Model for Satellite Imagerycited by 130
  12. 2023SkyScript: A Large and Semantically Diverse Vision-Language Dataset for Remote Sensingcited by 123
38 more
  1. 2023GEO-Bench: Toward Foundation Models for Earth Monitoringcited by 122
  2. 2023Towards Geospatial Foundation Models via Continual Pretrainingcited by 109
  3. 2023Lightweight, Pre-trained Transformers for Remote Sensing Timeseriescited by 102
  4. 2024Neural Plasticity-Inspired Multimodal Foundation Model for Earth Observationcited by 97
  5. 2024Rethinking Transformers Pre-training for Multi-Spectral Satellite Imagerycited by 82
  6. 2023UrbanCLIP: Learning Text-enhanced Urban Region Profiling with Contrastive Language-Image Pretraining from the Webcited by 77
  7. 2024Mission Critical -- Satellite Data is a Distinct Modality in Machine Learningcited by 76
  8. 2024MMEarth: Exploring Multi-Modal Pretext Tasks For Geospatial Representation Learningcited by 76
  9. 2024AnySat: One Earth Observation Model for Many Resolutions, Scales, and Modalitiescited by 73
  10. 2024HyperSIGMA: Hyperspectral Intelligence Comprehension Foundation Modelcited by 72
  11. 2023Remote Sensing Vision-Language Foundation Models without Annotations via Ground Remote Alignmentcited by 71
  12. 2024TEOChat: A Large Vision-Language Assistant for Temporal Earth Observation Datacited by 69
  13. 2023CSP: Self-Supervised Contrastive Spatial Pre-Training for Geospatial-Visual Representationscited by 67
  14. 2024MTP: Advancing Remote Sensing Foundation Model via Multi-Task Pretrainingcited by 66
  15. 2024OmniSat: Self-Supervised Modality Fusion for Earth Observationcited by 63
  16. 2023A Billion-scale Foundation Model for Remote Sensing Imagescited by 62
  17. 2024PANGAEA: A Global and Inclusive Benchmark for Geospatial Foundation Modelscited by 62
  18. 2025Towards a Unified Copernicus Foundation Model for Earth Visioncited by 52
  19. 2024Locate Anything on Earth: Advancing Open-Vocabulary Object Detection for Remote Sensing Communitycited by 50
  20. 2023SSL4EO-L: Datasets and Foundation Models for Landsat Imagerycited by 49
  21. 2023CMID: A Unified Self-Supervised Learning Framework for Remote Sensing Image Understandingcited by 48
  22. 2023Feature Guided Masked Autoencoder for Self-supervised Learning in Remote Sensingcited by 48
  23. 2024Cross-Scale MAE: A Tale of Multi-Scale Exploitation in Remote Sensingcited by 47
  24. 2024Bridging Remote Sensors with Multisensor Geospatial Foundation Modelscited by 46
  25. 2023Text2Seg: Remote Sensing Image Semantic Segmentation via Text-Guided Visual Foundation Modelscited by 44
  26. 2024SpectralEarth: Training Hyperspectral Foundation Models at Scalecited by 42
  27. 2024SatSynth: Augmenting Image-Mask Pairs through Diffusion Models for Aerial Semantic Segmentationcited by 41
  28. 2024MetaEarth: A Generative Foundation Model for Global-Scale Remote Sensing Image Generationcited by 41
  29. 2025Galileo: Learning Global & Local Features of Many Remote Sensing Modalitiescited by 41
  30. 2025RemoteSAM: Towards Segment Anything for Earth Observationcited by 38
  31. 2024Changen2: Multi-Temporal Remote Sensing Generative Change Foundation Modelcited by 37
  32. 2023Decoupling Common and Unique Representations for Multimodal Self-supervised Learningcited by 36
  33. 2025TESSERA: Temporal Embeddings of Surface Spectra for Earth Representation and Analysiscited by 35
  34. 2024One for All: Toward Unified Foundation Models for Earth Visioncited by 34
  35. 2025Panopticon: Advancing Any-Sensor Foundation Models for Earth Observationcited by 34
  36. 2023Channel Vision Transformers: An Image Is Worth 1 x 16 x 16 Wordscited by 33
  37. 2024Deep Learning for Cross-Domain Data Fusion in Urban Computing: Taxonomy, Advances, and Outlookcited by 33
  38. 2024GeoSynth: Contextually-Aware High-Resolution Satellite Image Synthesiscited by 32

50 most cited of 402 All 402 (JSON)