Resampling Augmentation for Time Series Contrastive Learning: Application to Remote Sensing

Antoine Saget, Baptiste Lafabregue, Antoine Cornuéjols, Pierre Gançarski
Proceedings of The TerraBytes {ICML} Workshop: Towards global datasets and models for Earth Observation, PMLR 292:32-45, 2025.

Abstract

Given the abundance of unlabeled Satellite Image Time Series (SITS) and the scarcity of labeled data, contrastive self-supervised pretraining emerges as a natural tool to leverage this vast quantity of unlabeled data. However, designing effective data augmentations for contrastive learning remains challenging for time series. We introduce a novel resampling-based augmentation strategy that generates positive pairs by temporally upsampling time series and extracting disjoint subsequences while preserving temporal coverage. We validate our approach on multiple agricultural classification benchmarks using Sentinel-2 imagery, showing that it outperforms common alternatives such as jittering, resizing, and masking. Further, we achieve state-of-the-art performance on the S2-Agri100 dataset without employing spatial information or temporal encodings, surpassing more complex mask-based SSL frameworks. Our method offers a simple, yet effective, contrastive learning augmentation for remote sensing time series.

Cite this Paper

BibTeX
@InProceedings{pmlr-v292-saget25a, title = {Resampling Augmentation for Time Series Contrastive Learning: Application to Remote Sensing}, author = {Saget, Antoine and Lafabregue, Baptiste and Cornu{\'e}jols, Antoine and Gan{\c{c}}arski, Pierre}, booktitle = {Proceedings of The TerraBytes {ICML} Workshop: Towards global datasets and models for Earth Observation}, pages = {32--45}, year = {2025}, editor = {Audebert, Nicolas and Azizpour, Hossein and Barrière, Valentin and Castillo Navarro, Javiera and Czerkawski, Mikolaj and Fang, Heng and Francis, Alistair and Marsocci, Valerio and Nascetti, Andrea and Yadav, Ritu}, volume = {292}, series = {Proceedings of Machine Learning Research}, month = {19 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v292/main/assets/saget25a/saget25a.pdf}, url = {https://proceedings.mlr.press/v292/saget25a.html}, abstract = {Given the abundance of unlabeled Satellite Image Time Series (SITS) and the scarcity of labeled data, contrastive self-supervised pretraining emerges as a natural tool to leverage this vast quantity of unlabeled data. However, designing effective data augmentations for contrastive learning remains challenging for time series. We introduce a novel resampling-based augmentation strategy that generates positive pairs by temporally upsampling time series and extracting disjoint subsequences while preserving temporal coverage. We validate our approach on multiple agricultural classification benchmarks using Sentinel-2 imagery, showing that it outperforms common alternatives such as jittering, resizing, and masking. Further, we achieve state-of-the-art performance on the S2-Agri100 dataset without employing spatial information or temporal encodings, surpassing more complex mask-based SSL frameworks. Our method offers a simple, yet effective, contrastive learning augmentation for remote sensing time series.} }
Endnote
%0 Conference Paper %T Resampling Augmentation for Time Series Contrastive Learning: Application to Remote Sensing %A Antoine Saget %A Baptiste Lafabregue %A Antoine Cornuéjols %A Pierre Gançarski %B Proceedings of The TerraBytes {ICML} Workshop: Towards global datasets and models for Earth Observation %C Proceedings of Machine Learning Research %D 2025 %E Nicolas Audebert %E Hossein Azizpour %E Valentin Barrière %E Javiera Castillo Navarro %E Mikolaj Czerkawski %E Heng Fang %E Alistair Francis %E Valerio Marsocci %E Andrea Nascetti %E Ritu Yadav %F pmlr-v292-saget25a %I PMLR %P 32--45 %U https://proceedings.mlr.press/v292/saget25a.html %V 292 %X Given the abundance of unlabeled Satellite Image Time Series (SITS) and the scarcity of labeled data, contrastive self-supervised pretraining emerges as a natural tool to leverage this vast quantity of unlabeled data. However, designing effective data augmentations for contrastive learning remains challenging for time series. We introduce a novel resampling-based augmentation strategy that generates positive pairs by temporally upsampling time series and extracting disjoint subsequences while preserving temporal coverage. We validate our approach on multiple agricultural classification benchmarks using Sentinel-2 imagery, showing that it outperforms common alternatives such as jittering, resizing, and masking. Further, we achieve state-of-the-art performance on the S2-Agri100 dataset without employing spatial information or temporal encodings, surpassing more complex mask-based SSL frameworks. Our method offers a simple, yet effective, contrastive learning augmentation for remote sensing time series.
APA
Saget, A., Lafabregue, B., Cornuéjols, A. & Gançarski, P.. (2025). Resampling Augmentation for Time Series Contrastive Learning: Application to Remote Sensing. Proceedings of The TerraBytes {ICML} Workshop: Towards global datasets and models for Earth Observation, in Proceedings of Machine Learning Research 292:32-45 Available from https://proceedings.mlr.press/v292/saget25a.html.

Related Material