MAESTRO: Masked AutoEncoders for Multimodal, Multitemporal, and Multispectral Earth Observation Data

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Labatie, Antoine, Vaccaro, Michael, Lardiere, Nina, Garioud, Anatol, Gonthier, Nicolas
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908584803041280
author Labatie, Antoine
Vaccaro, Michael
Lardiere, Nina
Garioud, Anatol
Gonthier, Nicolas
author_facet Labatie, Antoine
Vaccaro, Michael
Lardiere, Nina
Garioud, Anatol
Gonthier, Nicolas
contents Self-supervised learning holds great promise for remote sensing, but standard self-supervised methods must be adapted to the unique characteristics of Earth observation data. We take a step in this direction by conducting a comprehensive benchmark of fusion strategies and normalization schemes of reconstruction targets for multimodal, multitemporal, and multispectral Earth observation data. Based on our findings, we introduce MAESTRO, a novel adaptation of the Masked Autoencoder with optimized fusion mechanisms and a normalization scheme that incorporates a spectral prior as a self-supervisory signal. Evaluated on four Earth observation datasets in both intra- and cross-dataset settings, MAESTRO achieves state-of-the-art performance on tasks that strongly rely on multitemporal dynamics, while also remaining competitive on others. Code to reproduce all our experiments is available at https://github.com/ignf/maestro.
format Preprint
id arxiv_https___arxiv_org_abs_2508_10894
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle MAESTRO: Masked AutoEncoders for Multimodal, Multitemporal, and Multispectral Earth Observation Data
Labatie, Antoine
Vaccaro, Michael
Lardiere, Nina
Garioud, Anatol
Gonthier, Nicolas
Computer Vision and Pattern Recognition
Self-supervised learning holds great promise for remote sensing, but standard self-supervised methods must be adapted to the unique characteristics of Earth observation data. We take a step in this direction by conducting a comprehensive benchmark of fusion strategies and normalization schemes of reconstruction targets for multimodal, multitemporal, and multispectral Earth observation data. Based on our findings, we introduce MAESTRO, a novel adaptation of the Masked Autoencoder with optimized fusion mechanisms and a normalization scheme that incorporates a spectral prior as a self-supervisory signal. Evaluated on four Earth observation datasets in both intra- and cross-dataset settings, MAESTRO achieves state-of-the-art performance on tasks that strongly rely on multitemporal dynamics, while also remaining competitive on others. Code to reproduce all our experiments is available at https://github.com/ignf/maestro.
title MAESTRO: Masked AutoEncoders for Multimodal, Multitemporal, and Multispectral Earth Observation Data
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2508.10894