About

Carla Geara

PhD Candidate · Computer Vision and Remote Sensing · Kanop | Télécom Paris | ONERA

I am a PhD candidate at Kanop, in collaboration with Télécom Paris (LTCI) and ONERA (DTIS), working on synthetic aperture radar (SAR) remote sensing for forestry applications, with a focus on developing self-supervised despeckling algorithms and their application to forest monitoring.

A central motivation of my work is assessing the utility of Sentinel-1 C-band PolInSAR data, despite its more limited penetration in forest environments, given its wide spatial coverage, temporal availability, public accessibility, and high resolution.

Synthetic Aperture Radar PolInSAR Despeckling Forest Monitoring Deep Learning Sentinel-1 Interferometry Forest height

News

IEEE GRSS 2025 Letters Prize Paper Award

Our paper, Extending InSAR2InSAR to Sentinel-1 Data, has received the IEEE Geoscience and Remote Sensing Society 2025 Letters Prize Paper Award !

C. Geara, C. Gelas, L. d. Vitry, E. Colin and F. Tupin, "Extending InSAR2InSAR to Sentinel-1 Data," in IEEE Geoscience and Remote Sensing Letters, vol. 22, pp. 1-5, 2025, Art no. 4008105

This work builds on previous results from the remote sensing group at Télécom Paris.
Special thanks to Loïc Denis, Cristiano Ulundo-Mendes, and Emanuele Dalsasso, who paved the way with SAR2SAR and PolSAR2PolSAR, and to Roland Akiki and Jérémy Anger for preliminary discussions on Sentinel-1.
Code is available on RING.

DOI Code

Research projects

Self-supervised despeckling of multichannel SAR data

Development of a self-supervised despeckling algorithm for multichannel Sentinel-1 data, covering PolSAR, InSAR, and PolInSAR configurations. The method exploits the burst overlap in Sentinel-1 Interferometric Wide Swath data to apply the Noise2Noise self-supervised denoising paradigm.

Assessment of InSAR parameters for forest height estimation

Systematic evaluation of Sentinel-1 InSAR coherence and phase as proxies for forest height. The study examines coherence–height correlations across different forest sites, and assessing the practical limits of C-band PolInSAR for canopy height retrieval.

Publications

[ACCEPTED] Generalization of InSAR2InSAR to Sentinel-1 Multi-Channel Data: PolSAR and PolInSAR

Carla Geara, Colette Gelas, Louis De Vitry, Elise Colin, and Florence Tupin

EUSAR, 2026

code

Extending InSAR2InSAR to Sentinel-1 Data
Winner of the IEEE Geoscience and Remote Sensing Society 2025 Letters Prize Paper Award

Carla Geara, Colette Gelas, Louis De Vitry, Elise Colin, and Florence Tupin

IEEE Geoscience and Remote Sensing Letters, 2025

HAL DOI code

InSAR2InSAR: A Self-Supervised Method for InSAR Parameters Estimation

Carla Geara, Colette Gelas, Louis De Vitry, Elise Colin, and Florence Tupin

EUSIPCO, 2024

HAL DOI

Feature space data augmentation for viewpoint-robust action recognition in videos

Carla Geara, Aleksandr Setkov, Astrid Orcesi, Bertrand Luvison

ICIP, 2023

HAL DOI

Contact

Email carla.geara@kanop.io
Google Scholar Scholar profile
Affiliation Kanop · Télécom Paris (LTCI) · ONERA (DTIS)