Romain Thoreau

Total : 9 publications

2025

Communication dans un congrès

• Parameter-Efficient Adaptation of Geospatial Foundation Models through Embedding Deflection
  Romain Thoreau, Valerio Marsocci, Dawa Derksen
  <div><p>As large-scale heterogeneous data sets become increasingly available, adapting foundation models at low cost has become a key issue. Seminal w...
  hal-05589503 | ⬇️ Télécharger
  📚 Exporter : BibTeX | EndNote | DC
  
• Can multimodal representation learning by alignment preserve modality-specific information?
  Romain Thoreau, Jessie Levillain, Dawa Derksen
  <div><p>Combining multimodal data is a key issue in a wide range of machine learning tasks, including many remote sensing problems. In Earth observati...
  hal-05589525 | ⬇️ Télécharger
  📚 Exporter : BibTeX | EndNote | DC
  

Article dans une revue

• Physics-Informed Variational Autoencoders for Improved Robustness to Environmental Factors of Variation
  Romain Thoreau, Laurent Risser, Véronique Achard, Béatrice Berthelot, Xavier Briottet
  <div><p>The combination of machine learning models with physical models is a recent research path to learn robust data representations. In this paper,...
  hal-05589480 | ⬇️ Télécharger
  📚 Exporter : BibTeX | EndNote | DC
  

2024

Article dans une revue

• Toulouse Hyperspectral Data Set: A benchmark data set to assess semi-supervised spectral representation learning and pixel-wise classification techniques
  Romain Thoreau, Laurent Risser, Véronique Achard, Béatrice Berthelot, Xavier Briottet
  Airborne hyperspectral images can be used to map the land cover in large urban areas, thanks to their very high spatial and spectral resolutions on a ...
  hal-04782619 | ⬇️ Télécharger
  📚 Exporter : BibTeX | EndNote | DC
  

2023

Thèse

• Semantic segmentation of airborne hyperspectral images (0.4 - 2.5 µm) for mapping impermeable surfaces in large urban areas
  Romain Thoreau
  In Earth observation, one of the main challenges is to map soil artificialization, which was identified as a major public policy issue in the 2018 Bio...
  tel-04660690 | ⬇️ Télécharger
  📚 Exporter : BibTeX | EndNote | DC
  

2022

Communication dans un congrès

• Probabilistic Breaking Tie: An Active Learning Strategy To Leverage Class Hierarchy For Impervious Surfaces Classification
  Romain Thoreau, Véronique Achard, Laurent Risser, Béatrice Berthelot, Xavier Briottet
  Urban sprawl has significantly changed the land cover of metropolis and periurban areas. In that regard, the monitoring of impervious surfaces has bec...
  hal-04133665 | ⬇️ Télécharger
  📚 Exporter : BibTeX | EndNote | DC
  
• Active learning on large hyperspectral datasets: a preprocessing method
  R. Thoreau, Véronique Achard, L. Risser, B. Berthelot, Xavier Briottet
  Machine learning algorithms demonstrated promising results for hyperspectral semantic segmentation. However, they strongly rely on the quality of trai...
  hal-03700279 | ⬇️ Télécharger
  📚 Exporter : BibTeX | EndNote | DC
  

Article dans une revue

• Active Learning for Hyperspectral Image Classification: A Comparative Review
  Romain Thoreau, Veronique Achard, Laurent Risser, Beatrice Berthelot, Xavier Briottet
  Machine learning algorithms have demonstrated impressive results for land cover mapping from hyperspectral data. To enhance the generalization capabil...
  hal-03676575 | ⬇️ Télécharger
  📚 Exporter : BibTeX | EndNote | DC
  

2021

Communication dans un congrès

• Hyperspectral Classification Based on Spectral Indices Learned Through Soft Attention Units
  Romain Thoreau, Veronique Achard, Xavier Briottet
  Classification of hyperspectral scenes has been dominated in recent years by Convolutional Neural Networks (CNNs). Spectral and spatial convolutions h...
  hal-03417120 | ⬇️ Télécharger
  📚 Exporter : BibTeX | EndNote | DC