Research output

Publications

Peer-reviewed work spanning production-oriented sequential user modeling and earlier research in deep learning for satellite SAR time series.

Industry machine learning

  1. 2026

    Abacus: Self-Supervised Event Counting-Aligned Distributional Pretraining for Sequential User Modeling

    Sullivan Castro, Artem Betlei, Thomas Di Martino, Nadir El Manouzi

    ACM Web Search and Data Mining (WSDM), pp. 1083–1088

Deep learning and Earth observation

  1. 2024

    Convolutional Autoencoder Applied to Short SAR Time Series for Under Canopy Object Detection

    Thomas Di Martino, Régis Guinvarc'h, Laetitia Thirion-Lefevre, Élise Colin

    IEEE IGARSS, pp. 2785–2788

  2. 2024

    Unsupervised Deep Learning for Vegetation Monitoring Using C-Band SAR Time Series: From Agriculture to Boreal Forests

    Thomas Di Martino

    PhD thesis, Université Paris-Saclay

  3. 2023

    Towards the Understanding of the C-Band Temporal Signature of Boreal Forest Through Physiology Parameters Retrieval from Sentinel-1 Time Series and Machine Learning

    Thomas Di Martino, Régis Guinvarc'h, Laetitia Thirion-Lefevre, Élise Colin

    IEEE IGARSS, pp. 5551–5554

  4. 2023

    Detection of Forest Fires through Deep Unsupervised Learning Modeling of Sentinel-1 Time Series

    Thomas Di Martino, Bertrand Le Saux, Régis Guinvarc'h, Laetitia Thirion-Lefevre, Élise Colin

    ISPRS International Journal of Geo-Information, 12(8), 332

  5. 2023

    Grad-SLAM: Explaining Convolutional Autoencoders' Latent Space of Satellite Image Time Series

    Thomas Di Martino, Régis Guinvarc'h, Laetitia Thirion-Lefevre, Élise Colin

    IEEE Geoscience and Remote Sensing Letters

  6. 2022

    FARMSAR: Fixing Agricultural Mislabels Using Sentinel-1 Time Series and Autoencoders

    Thomas Di Martino, Régis Guinvarc'h, Laetitia Thirion-Lefevre, Élise Colin

    Remote Sensing, 15(1), 35

  7. 2022

    Modelling of Agricultural SAR Time Series Using Convolutional Autoencoder for the Extraction of Harvesting Practices of Rice Fields

    Thomas Di Martino, Élise Colin, Laetitia Thirion-Lefevre, Régis Guinvarc'h

    14th European Conference on Synthetic Aperture Radar (EUSAR)

  8. 2022

    Extracting Relevance from SAR Temporal Profiles on a Glacier and an Alpine Watershed by a Deep Autoencoder

    Laurane Charrier, Thomas Di Martino, Élise Colin Koeniguer, Flora Weissgerber, Aurélien Plyer

    International Archives of Photogrammetry, Remote Sensing and Spatial Information Sciences, 43, 1309–1316

  9. 2021

    Beets or Cotton? Blind Extraction of Fine Agricultural Classes Using a Convolutional Autoencoder Applied to Temporal SAR Signatures

    Thomas Di Martino, Régis Guinvarc'h, Laetitia Thirion-Lefevre, Élise Colin Koeniguer

    IEEE Transactions on Geoscience and Remote Sensing

  10. 2021

    Convolutional Autoencoder for Unsupervised Representation Learning of PolSAR Time-Series

    Thomas Di Martino, Régis Guinvarc'h, Laetitia Thirion-Lefevre, Élise Colin Koeniguer

    IEEE IGARSS

  11. 2021

    Multi-Branch Deep Learning Model for Detection of Settlements Without Electricity

    Thomas Di Martino, Maxime Lenormand, Élise Colin Koeniguer

    IEEE IGARSS

Earlier work

  1. 2020

    REACTIV Algorithm

    Thomas Di Martino, Élise Colin-Koeniguer, Régis Guinvarc'h, Laetitia Thirion-Lefevre

    Sentinel Hub Custom Script Competition

  2. 2020

    Multimodal Similarity Learning for Duplicate Product Identification

    Thomas Di Martino

    MSc thesis, Heriot-Watt University