Remote Sensing Lab, National Technical University of Athens

The Remote Sensing Lab at NTUA specializes in advanced methodologies for analyzing multi-modal and multi-temporal Earth Observation (EO) data, integrating cutting-edge machine learning and AI techniques to address pressing environmental and societal challenges. Our research spans atmospheric, radiometric, and geometric corrections, object-based data analysis, and feature extraction for object detection in both 2D and 3D. We work with diverse remote sensing platforms, including satellites, drones, and gliders, leveraging multispectral, hyperspectral, thermal, and radar imaging to support applications in forestry, agriculture, hydrology, disaster monitoring, land and water resource management, environmental pollution, and infrastructure studies. Equipped with state-of-the-art IT infrastructure, including multiple 3090 RTX GPUs and 200 TB of machine-learning-ready datasets, our lab enables rapid prototyping, real-time processing, and large-scale data analytics. Our research is driven by a passion for harnessing big EO data and machine learning to generate meaningful insights for environmental sustainability and societal well-being. We focus on remote sensing foundation models, self-supervised learning for multi-modal EO data, and vision-language models for remote sensing interpretation. We participate in international research programs (e.g. H2020 RAMONES, H2020 iTOBOS). Our work also extends to earth system deep learning for spatiotemporal forecasting, supporting AI4EO applications for climate change impact assessment and disaster management. We actively coordinate research initiatives under EUSPA, H2020, Horizon Europe and ESA, such as projects ThinkingEarth, MeDiTwin, DeepCube, and SeasFire, pushing the frontiers of AI-driven EO analytics to enhance scientific understanding and decision-making.