Information
📚 Organization | University of East London |
📆 Period @ ESA Φ-lab | January 2026 - June 2026 |
🌍 Project @ ESA CIN | ⚠️ Project is ongoing - Foundation Models Benchmarking |
📍 GitHub | |
🔨 Linkedin |
Bio
Kenzo Bounegta is a researcher specialising in Deep Learning and Computer Vision for satellite imagery. He holds a Master’s degree in Artificial Intelligence from École Polytechnique and HEC Paris. His work has focused on applications such as flood detection, land-use classification, and the detection of looted archaeological sites in Afghanistan, using both optical and SAR satellite data. He has a strong interest in multimodal Geo-Foundation Models.
Current Role
He is joining ESA Φ-lab as a Visiting Researcher, where his work focuses on benchmarking Foundation Models for Earth Observation. His activities include fine-tuning experiments on datasets using TerraMind, Copernicus Foundation Models, and Galileo architectures. The outcomes of this work include the integration of new datasets and models into PANGAEA, as well as the publication of benchmark results for satellite imagery.
Areas of Expertise
Kenzo works primarily with PyTorch on tasks such as semantic segmentation, crop classification, and infrastructure detection from satellite imagery. His technical expertise centres on Foundation Models, Vision–Language Models, and computer vision techniques applied to multi-sensor Earth Observation data.
Vision for the Future
His goal is to develop reproducible benchmarks for Foundation Models that can scale to operational Earth Observation systems, supporting applications in climate monitoring and disaster response.