Skip to main content
Springer Nature Link
Log in

MMEarth: Exploring Multi-modal Pretext Tasks for Geospatial Representation Learning

  • Conference paper
  • First Online:
Computer Vision – ECCV 2024 (ECCV 2024)

Part of the book series: Lecture Notes in Computer Science ((LNCS,volume 15122))

Included in the following conference series:

  • 313 Accesses

Abstract

The volume of unlabelled Earth observation (EO) data is huge, but many important applications lack labelled training data. However, EO data offers the unique opportunity to pair data from different modalities and sensors automatically based on geographic location and time, at virtually no human labor cost. We seize this opportunity to create MMEarth, a diverse multi-modal pretraining dataset at global scale. Using this new corpus of 1.2 million locations, we propose a Multi-Pretext Masked Autoencoder (MP-MAE) approach to learn general-purpose representations for optical satellite images. Our approach builds on the ConvNeXt V2 architecture, a fully convolutional masked autoencoder (MAE). Drawing upon a suite of multi-modal pretext tasks, we demonstrate that our MP-MAE approach outperforms both MAEs pretrained on ImageNet and MAEs pretrained on domain-specific satellite images. This is shown on several downstream tasks including image classification and semantic segmentation. We find that pretraining with multi-modal pretext tasks notably improves the linear probing performance compared to pretraining on optical satellite images only. This also leads to better label efficiency and parameter efficiency which are crucial aspects in global scale applications. (The MMEarth dataset is available on the project page: vishalned.github.io/mmearth. The dataset construction code is available here: github.com/vishalned/MMEarth-data. The MP-MAE code for training and evaluation is available here: github.com/vishalned/MMEarth-train).

This is a preview of subscription content, log in via an institution to check access.

Access this chapter

Log in via an institution

Subscribe and save

Springer+ Basic
$34.99 /Month
  • Get 10 units per month
  • Download Article/Chapter or eBook
  • 1 Unit = 1 Article or 1 Chapter
  • Cancel anytime
Subscribe now

Buy Now

Chapter
USD 29.95
Price excludes VAT (USA)
  • Available as PDF
  • Read on any device
  • Instant download
  • Own it forever

Tax calculation will be finalised at checkout

Purchases are for personal use only

Institutional subscriptions

Similar content being viewed by others

Notes

  1. 1.

    We use the latest GEO-Bench version v1.0 in which datasets were class-balanced.

References

  1. Argaw, D.M., Lee, J.Y., Woodson, M., Kweon, I.S., Caba Heilbron, F.: Long-range multimodal pretraining for movie understanding. In: International Conference on Computer Vision (ICCV). IEEE (2023)

    Google Scholar 

  2. Assran, M., et al.: Self-supervised learning from images with a joint-embedding predictive architecture. In: Computer Vision and Pattern Recognition (CVPR), pp. 15619–15629 (2023)

    Google Scholar 

  3. Ayush, K., et al.: Geography-aware self-supervised learning. In: International Conference on Computer Vision (ICCV), pp. 10181–10190 (2021)

    Google Scholar 

  4. Bachmann, R., Mizrahi, D., Atanov, A., Zamir, A.: MultiMAE: multi-modal multi-task masked autoencoders. In: Avidan, S., Brostow, G., Cisse, M., Farinella, G.M., Hassner, T. (eds.) European Conference on Computer Vision (ECCV), pp. 348–367. Springer, Cham (2022). https://doi.org/10.1007/978-3-031-19836-6_20

  5. Balestriero, R., et al.: A cookbook of self-supervised learning. arXiv preprint arXiv:2304.12210 (2023)

  6. Bardes, A., et al.: Revisiting feature prediction for learning visual representations from video. arXiv preprint (2024)

    Google Scholar 

  7. Bastani, F., Wolters, P., Gupta, R., Ferdinando, J., Kembhavi, A.: SatlasPretrain: a large-scale dataset for remote sensing image understanding. In: International Conference on Computer Vision (ICCV), pp. 16772–16782 (2023)

    Google Scholar 

  8. Brown, C.F., et al.: Dynamic world, near real-time global 10 m land use land cover mapping. Sci. Data 9(1), 251 (2022)

    Article  Google Scholar 

  9. Choy, C., Gwak, J., Savarese, S.: 4D spatio-temporal convnets: Minkowski convolutional neural networks. In: Computer Vision and Pattern Recognition (CVPR), pp. 3075–3084 (2019)

    Google Scholar 

  10. Christie, G., Fendley, N., Wilson, J., Mukherjee, R.: Functional map of the world. In: Computer Vision and Pattern Recognition (CVPR), pp. 6172–6180 (2018)

    Google Scholar 

  11. Cong, Y., et al.: SatMAE: pre-training transformers for temporal and multi-spectral satellite imagery. Adv. Neural Inf. Process. Syst. (NeurIPS) 35, 197–211 (2022)

    Google Scholar 

  12. Daudt, R.C., Wulf, H., Hafner, E.D., Bühler, Y., Schindler, K., Wegner, J.D.: Snow depth estimation at country-scale with high spatial and temporal resolution. ISPRS J. Photogramm. Remote. Sens. 197, 105–121 (2023)

    Article  Google Scholar 

  13. Deng, J., Dong, W., Socher, R., Li, L.J., Li, K., Fei-Fei, L.: ImageNet: a large-scale hierarchical image database. In: Computer Vision and Pattern Recognition (CVPR), pp. 248–255. IEEE (2009)

    Google Scholar 

  14. Devlin, J., Chang, M.W., Lee, K., Toutanova, K.: BERT: pre-training of deep bidirectional transformers for language understanding. In: Burstein, J., Doran, C., Solorio, T. (eds.) Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, vol. 1 (Long and Short Papers), pp. 4171–4186. Association for Computational Linguistics, Minneapolis, Minnesota (2019)

    Google Scholar 

  15. Dinerstein, E., et al.: An ecoregion-based approach to protecting half the terrestrial realm. Bioscience 67(6), 534–545 (2017)

    Article  Google Scholar 

  16. Dosovitskiy, A., et al.: An image is worth 16 \(\times \) 16 words: transformers for image recognition at scale. In: International Conference on Learning Representations (ICLR) (2021)

    Google Scholar 

  17. Dubayah, R., et al.: The global ecosystem dynamics investigation: high-resolution laser ranging of the earth’s forests and topography. Sci. Remote Sens. 1, 100002 (2020)

    Article  Google Scholar 

  18. Feichtenhofer, C., Fan, H., Li, Y., He, K.: Masked autoencoders as spatiotemporal learners. In: Oh, A.H., Agarwal, A., Belgrave, D., Cho, K. (eds.) Advances in Neural Information Processing Systems (NeurIPS) (2022)

    Google Scholar 

  19. Geng, X., Liu, H., Lee, L., Schuurmans, D., Levine, S., Abbeel, P.: Multimodal masked autoencoders learn transferable representations (2022)

    Google Scholar 

  20. Ghiasi, G., Zoph, B., Cubuk, E.D., Le, Q.V., Lin, T.Y.: Multi-task self-training for learning general representations. In: International Conference on Computer Vision (ICCV), pp. 8856–8865 (2021)

    Google Scholar 

  21. Gorelick, N., Hancher, M., Dixon, M., Ilyushchenko, S., Thau, D., Moore, R.: Google earth engine: planetary-scale geospatial analysis for everyone. Remote Sens. Environ. 202, 18–27 (2017)

    Article  Google Scholar 

  22. He, K., Chen, X., Xie, S., Li, Y., Dollár, P., Girshick, R.: Masked autoencoders are scalable vision learners. In: Computer Vision and Pattern Recognition (CVPR), pp. 16000–16009 (2022)

    Google Scholar 

  23. Helber, P., Bischke, B., Dengel, A., Borth, D.: EuroSAT: a novel dataset and deep learning benchmark for land use and land cover classification. IEEE J. Sel. Topics Appl. Earth Observ. Remote Sens. 12(7), 2217–2226 (2019)

    Article  Google Scholar 

  24. Kendall, A., Gal, Y., Cipolla, R.: Multi-task learning using uncertainty to weigh losses for scene geometry and semantics. In: Computer Vision and Pattern Recognition (CVPR), pp. 7482–7491 (2018)

    Google Scholar 

  25. Klemmer, K., Rolf, E., Robinson, C., Mackey, L., Rußwurm, M.: SatCLIP: global, general-purpose location embeddings with satellite imagery. arXiv preprint arXiv:2311.17179 (2023)

  26. Kruitwagen, L., Story, K., Friedrich, J., Byers, L., Skillman, S., Hepburn, C.: A global inventory of photovoltaic solar energy generating units. Nature 598(7882), 604–610 (2021)

    Article  Google Scholar 

  27. Lacoste, A., et al.: GEO-Bench: toward foundation models for earth monitoring. In: Advances in Neural Information Processing Systems (NeurIPS) Datasets and Benchmarks Track (2023)

    Google Scholar 

  28. Lang, N., Jetz, W., Schindler, K., Wegner, J.D.: A high-resolution canopy height model of the earth. Nat. Ecol. Evol. 7(11), 1778–1789 (2023)

    Article  Google Scholar 

  29. Manas, O., Lacoste, A., Giró-i Nieto, X., Vazquez, D., Rodriguez, P.: Seasonal contrast: unsupervised pre-training from uncurated remote sensing data. In: International Conference on Computer Vision (ICCV), pp. 9414–9423 (2021)

    Google Scholar 

  30. Microsoft Open Source, McFarland, M., Emanuele, R., Morris, D., Augspurger, T.: microsoft/planetarycomputer, October 2022. https://doi.org/10.5281/zenodo.7261897

  31. Mizrahi, D., et al.: 4M: massively multimodal masked modeling. In: Advances in Neural Information Processing Systems (NeurIPS) (2023)

    Google Scholar 

  32. Mohamed, A., et al.: Self-supervised speech representation learning: a review. IEEE J. Sel. Topics Signal Process. 16(6), 1179–1210 (2022)

    Article  Google Scholar 

  33. Mommert, M., Kesseli, N., Hanna, J., Scheibenreif, L., Borth, D., Demir, B.: Ben-Ge: extending BigEarthNet with geographical and environmental data. In: IGARSS 2023-2023 IEEE International Geoscience and Remote Sensing Symposium, pp. 1016–1019. IEEE (2023)

    Google Scholar 

  34. Oquab, M., et al.: DINOv2: learning robust visual features without supervision (2023)

    Google Scholar 

  35. Pathak, D., Krahenbuhl, P., Donahue, J., Darrell, T., Efros, A.A.: Context encoders: feature learning by inpainting. In: Computer Vision and Pattern Recognition (CVPR), pp. 2536–2544 (2016)

    Google Scholar 

  36. Planet, Radiant Earth Foundation, Western Cape Department of Agriculture, German Aerospace Center (DLR): A fusion dataset for crop type classification in Western Cape, South Africa (2021). https://doi.org/10.34911/RDNT.GQY868

  37. Radford, A., et al.: Learning transferable visual models from natural language supervision. In: International Conference on Machine Learning (ICML), pp. 8748–8763. PMLR (2021)

    Google Scholar 

  38. Reed, C.J., et al.: Scale-MAE: a scale-aware masked autoencoder for multiscale geospatial representation learning. In: International Conference on Computer Vision (ICCV), pp. 4088–4099 (2023)

    Google Scholar 

  39. Ronneberger, O., Fischer, P., Brox, T.: U-Net: convolutional networks for biomedical image segmentation. In: Navab, N., Hornegger, J., Wells, W., Frangi, A. (eds.) Medical Image Computing and Computer-Assisted Intervention (MICCAI), pp. 234–241. Springer, Cham (2015). https://doi.org/10.1007/978-3-319-24574-4_28

  40. Rußwurm, M., Venkatesa, S.J., Tuia, D.: Large-scale detection of marine debris in coastal areas with Sentinel-2. IScience 26(12), 108402 (2023)

    Google Scholar 

  41. de Sa, V.R., Ballard, D.H.: Category learning through multimodality sensing. Neural Comput. 10(5), 1097–1117 (1998)

    Article  Google Scholar 

  42. Sumbul, G., Charfuelan, M., Demir, B., Markl, V.: BigEarthNet: a large-scale benchmark archive for remote sensing image understanding. In: IGARSS 2019-2019 IEEE International Geoscience and Remote Sensing Symposium, pp. 5901–5904. IEEE (2019)

    Google Scholar 

  43. Sumbul, G., et al.: BigEarthNet-MM: a large-scale, multimodal, multilabel benchmark archive for remote sensing image classification and retrieval [software and data sets]. IEEE Geosci. Remote Sens. Mag. 9(3), 174–180 (2021)

    Article  Google Scholar 

  44. Tolan, J., et al.: Very high resolution canopy height maps from RGB imagery using self-supervised vision transformer and convolutional decoder trained on aerial lidar. Remote Sens. Environ. 300, 113888 (2024)

    Article  Google Scholar 

  45. Tong, Z., Song, Y., Wang, J., Wang, L.: VideoMAE: masked autoencoders are data-efficient learners for self-supervised video pre-training. In: Oh, A.H., Agarwal, A., Belgrave, D., Cho, K. (eds.) Advances in Neural Information Processing Systems (NeurIPS) (2022)

    Google Scholar 

  46. Tseng, G., Zvonkov, I., Purohit, M., Rolnick, D., Kerner, H.: Lightweight, pre-trained transformers for remote sensing timeseries. arXiv preprint arXiv:2304.14065 (2023)

  47. Tucker, C., et al.: Sub-continental-scale carbon stocks of individual trees in African drylands. Nature 615(7950), 80–86 (2023)

    Article  Google Scholar 

  48. Van Horn, G., Cole, E., Beery, S., Wilber, K., Belongie, S., Mac Aodha, O.: Benchmarking representation learning for natural world image collections. In: Computer Vision and Pattern Recognition (CVPR), pp. 12884–12893 (2021)

    Google Scholar 

  49. Vandenhende, S., Georgoulis, S., Van Gansbeke, W., Proesmans, M., Dai, D., Van Gool, L.: Multi-task learning for dense prediction tasks: a survey. IEEE Trans. Pattern Anal. Mach. Intell. 44(7), 3614–3633 (2021)

    Google Scholar 

  50. Vincent, P., Larochelle, H., Bengio, Y., Manzagol, P.A.: Extracting and composing robust features with denoising autoencoders. In: International Conference on Machine Learning (ICML), pp. 1096–1103. ACM (2008)

    Google Scholar 

  51. Wang, Y., Braham, N.A.A., Xiong, Z., Liu, C., Albrecht, C.M., Zhu, X.X.: SSL4EO-S12: a large-scale multi-modal, multi-temporal dataset for self-supervised learning in earth observation. arXiv preprint arXiv:2211.07044 (2022)

  52. Wei, C., Fan, H., Xie, S., Wu, C.Y., Yuille, A., Feichtenhofer, C.: Masked feature prediction for self-supervised visual pre-training. In: Computer Vision and Pattern Recognition (CVPR), pp. 14668–14678 (2022)

    Google Scholar 

  53. Woo, S., et al.: ConvNeXt V2: co-designing and scaling convnets with masked autoencoders. In: Computer Vision and Pattern Recognition (CVPR), pp. 16133–16142 (2023)

    Google Scholar 

  54. Xie, Z., et al.: SimMIM: a simple framework for masked image modeling. In: Computer Vision and Pattern Recognition (CVPR), pp. 9653–9663 (2022)

    Google Scholar 

  55. Yin, L., et al.: Mapping smallholder cashew plantations to inform sustainable tree crop expansion in Benin. Remote Sens. Environ. 295, 113695 (2023)

    Article  Google Scholar 

  56. Yu, X., Tang, L., Rao, Y., Huang, T., Zhou, J., Lu, J.: Point-BERT: pre-training 3D point cloud transformers with masked point modeling. In: Computer Vision and Pattern Recognition (CVPR), pp. 19313–19322 (2022)

    Google Scholar 

  57. Zamir, A.R., Sax, A., Shen, W., Guibas, L.J., Malik, J., Savarese, S.: Taskonomy: disentangling task transfer learning. In: Computer Vision and Pattern Recognition (CVPR), pp. 3712–3722 (2018)

    Google Scholar 

  58. Zhu, X.X., et al.: So2Sat LCZ42: a benchmark data set for the classification of global local climate zones [software and data sets]. IEEE Geosci. Remote Sens. Mag. 8(3), 76–89 (2020)

    Article  MathSciNet  Google Scholar 

Download references

Acknowledgments

We thank Lucia Gordon for the valuable feedback. We greatly appreciate the open data policies of the Copernicus program and its partners ESA and ECMWF. We thank Google Earth Engine for hosting the data and providing free access. This work was supported in part by the Pioneer Centre for AI, DNRF grant number P1. The authors AK, CI, and NL acknowledge support by the research grant DeReEco (grant number 34306) from Villum Foundation. SO and CI acknowledge support by the research grant Global Wetland Center (grant number NNF23OC0081089) from Novo Nordisk Foundation. CI and SB acknowledge support by the European Union project ELIAS (grant agreement number 101120237). We thank the Danish e-Infrastructure Consortium (DeiC), Martin Brandt, and Konrad Schindler for their support with computing resources.

Author information

Authors and Affiliations

  1. University of Copenhagen, Copenhagen, Denmark

    Vishal Nedungadi, Ankit Kariryaa, Stefan Oehmcke, Serge Belongie, Christian Igel & Nico Lang

Authors
  1. Vishal Nedungadi
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Ankit Kariryaa
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Stefan Oehmcke
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Serge Belongie
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Christian Igel
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Nico Lang
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Nico Lang .

Editor information

Editors and Affiliations

  1. University of Birmingham, Birmingham, UK

    Aleš Leonardis

  2. University of Trento, Trento, Italy

    Elisa Ricci

  3. Technical University of Darmstadt, Darmstadt, Germany

    Stefan Roth

  4. Princeton University, Princeton, NJ, USA

    Olga Russakovsky

  5. Czech Technical University in Prague, Prague, Czech Republic

    Torsten Sattler

  6. École des Ponts ParisTech, Marne-la-Vallée, France

    Gül Varol

1 Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary material 1 (pdf 1183 KB)

Rights and permissions

Reprints and permissions

Copyright information

© 2025 The Author(s), under exclusive license to Springer Nature Switzerland AG

About this paper

Check for updates. Verify currency and authenticity via CrossMark

Cite this paper

Nedungadi, V., Kariryaa, A., Oehmcke, S., Belongie, S., Igel, C., Lang, N. (2025). MMEarth: Exploring Multi-modal Pretext Tasks for Geospatial Representation Learning. In: Leonardis, A., Ricci, E., Roth, S., Russakovsky, O., Sattler, T., Varol, G. (eds) Computer Vision – ECCV 2024. ECCV 2024. Lecture Notes in Computer Science, vol 15122. Springer, Cham. https://doi.org/10.1007/978-3-031-73039-9_10

Download citation

  • DOI: https://doi.org/10.1007/978-3-031-73039-9_10

  • Published:

  • Publisher Name: Springer, Cham

  • Print ISBN: 978-3-031-73038-2

  • Online ISBN: 978-3-031-73039-9

  • eBook Packages: Computer ScienceComputer Science (R0)

Publish with us

Policies and ethics