Next Tuesday 21st November, ESE’s DataLearning group will be hosting a special session for their Seminar series with the phi-lab fellows of the European Space Agency (ESA). On this occasion, the session will include multiple talks and will start at 3 pm (GMT).
About the session:
Nikolaos Dionelis will speak about ‘the Earth Observation Foundation Model and the impact of pre-training tasks on downstream applications’
I am a Research Fellow at the European Space Agency (ESA). My background is in deep learning and computer vision. I have experience in (i) deep discriminative classifier models, including Convolutional Neural Networks (CNN), Residual Networks (ResNet), and Transformer models, and (ii) deep generative models, including Generative Adversarial Networks (GAN), invertible flow-based models, Variational Autoencoders (VAE), and diffusion models. I am enthusiastic about collaborating in projects to contribute to Earth Observation (EO) science. I was at Imperial College London for 8 years, i.e. from 2011 to 2019, PhD degree in Signal Processing and a Masters MEng degree (including Bachelor’s level study) in Electrical and Electronic Engineering.
Alistair Francis will speak about ‘Sensor Independent Cloud Masking with Hyperspectral Data’
Alistair Francis completed his PhD at Mullard Space Science Laboratory, UCL, working on improving deep learning’s generalisability to different sensors and modalities, with applications in both Earth Observation and Martian science. Since then, as a research fellow at the phi-lab, he has continued working on problems relating to deep learning in remote sensing, with projects on topics such as cloud masking, land cover classification, and an ongoing interest in representation learning with heterogeneous, multimodal sources.
Quentin Paletta will speak about ‘Solar Energy Forecasting from Cloud Cover Observations’
Quentin Paletta is an ESA research fellow in AI and the use of Earth Observations for Climate. His research focuses on predicting the impact of weather on solar power production at different spatio-temporal scales (local and regional, from minutes to hours in the future). This aims at accelerating the transition towards renewable energy sources by addressing the inherent variability of solar energy caused by meteorological influences such as aerosols and clouds. More specifically, his work involves developing short-term solar forecasting methods based on cloud cover observations from geostationary satellites or ground-level sky cameras. His research interests include power systems, weather forecasting, remote sensing, machine learning, and image and video analysis. In 2023, Quentin obtained his PhD from the Engineering Department of the University of Cambridge (UK). His thesis on solar energy meteorology was conducted in collaboration with ENGIE lab CRIGEN (France). Previously, he received an MSc from Centrale Supélec engineering school (Paris-Saclay University, France) and an MPhil in Energy Technologies from the University of Cambridge.
Alessandro Sebastianelli will speak about ‘Quantum Computing for Earth Observation’
Alessandro Sebastianelli received a degree in electronic engineering for automation and telecommunications from the University of Sannio, Benevento, Italy, in 2019, where he also pursued a PhD degree. His research topics mainly focus on remote sensing and satellite data analysis, artificial intelligence (AI) techniques for Earth observation, data fusion and quantum machine learning. He has coauthored several articles in reputed journals and conferences for the sector of remote sensing. Ha has been a Visited Researcher with Φ-lab, European Space Agency ESA/European Space Research Institute ESRIN, Frascati, Italy. He has won an ESA OSIP proposal in August 2020. He received an IEEE award for one of the best thesis in geoscience and remote sensing. Currently, he works as a Research Fellow in Quantum Computing for Earth Observation at ESA Φ-lab.
Mikolaj Czerkawski will speak about ‘Generative AI for Earth Observation’
Mikolaj Czerkawski is an Internal Research Fellow at ESA Φ-lab. He received his BEng and PhD from the University of Strathclyde in Glasgow, Scotland, specialising in applications of computer vision to Earth observation data. His research interests include image synthesis, generative models, and use cases involving restoration tasks of satellite imagery.
Peter Naylor will speak about ‘Point cloud analysis for Earth Observation’
Peter Naylor is an Internal Research Fellow at ESA Φ-lab. He obtained a PhD in Bioinformatics from Mines ParisTech in Paris, France, specialising in applications of computer vision to huge image data. His research interests include large image analysis, in particular, weekly supervised models. Lately, he has been interested in the analysis of point cloud data with physics-informed neural networks(PINNS)with applications in culture heritage, ocean topography and the Ice Sheet topography in Greenland.