My main focus within the Φ-lab is on the application of computer vision techniques in Earth Observation sci- ence. I have proposed and now lead three case stud- ies: (1) crop types mapping using drones, Copernicus Sentinel-2 and transfer learning from computer vision (in collaboration with World Food Program, UNICEF and Stanford University), (2) ML analysis of Swarm data (in collaboration with Istituto Nazionale di Geofisica e Vulcanologia and ESA Swarm team) and (3) “Seeing through the clouds” com- petition (in collaboration with CLAIRE and ESA Advance Concept Team).
My background is in computer science with a MSc and PhD obtained at Warsaw University of Technology. I started my professional research ana- lysing computer vision images in the field of pattern recognition and artifi- cial intelligence. Then, I enriched my experience with analysis of biomedical images oriented on stroke tissue recognition and renal cancer detection. In parallel, I have further developed my interests in EO optical data since join- ing the Space Research Centre of the Polish Academy of Science in 2011. I have developed new approaches for land cover classifications, features extraction and time series modelling for crop yield prediction within the projects funded by ESA, EU, national institutions and industry.
As an entrepreneurial and curious grad student, I believe remote sensing has an increasingly important role in helping to address environmental problems. The availability of a wealth of free and open data from space provides a useful tool to help define resilience and adaptation strategies and monitor their effectiveness. In the Φ-lab, I investigated the application of AI on multi-temporal and multi-sensor (Copernicus Sentinel-1 and Sentinel-2) images over rural areas for crop mapping, with Google Earth Engine. Crop mapping helps land use planning monitoring, reducing food waste and enhancing water management. I had the chance to meet inspirational professionals and experts. From May 2019, I spent six months in the team improving my knowledge in the field and working on my master thesis project.
My background is Environmental Science. I have a degree in structure and dynamics of the atmosphere and a MSc from University of Bologna (Italy) with EIT Climate-KIC certificate. I have been working on developing my start-up and, thanks to this opportunity, I gave a new input to my business idea as well as sharing remote sensing capabilities with other researchers and decision-makers with whom I am collaborating.
My passion is the study of our planet and its evolution due to climate change. In particular, the focus of my research is the use of InSAR methods to better understand the fast changing areas of the cryosphere. I am fascinated by the Machine Learning potential in these fields and interested in evaluating how AI tools can improve and facilitate the comprehension of what we can observe today through
satellite imagery. This interest grew from my earlier work at Φ-lab in 2018-2019, where I was involved in testing Machine Learning
algorithms and in designing Earth Observation datasets for Artificial Intelligence exploitation. The goal was to develop a series of innovation events, including Hackathons, to foster the use of specific Airborne and Satellite data.
My background is in Aerospace Engineering, and I have focused on EO and geoscience since I started my first internship at the Italian National Institute for Nuclear Physics (INFN) in 2017. While pursuing my academic degree at University of Rome La Sapienza, I had the possibility to engage in two other internships, respectively at ESA ESRIN (Φ-lab) and NASA JPL, where I have been working on different Earth scenarios, from the Poles to the Tropics,
focusing on SAR applications and performing environmental monitoring.
My new visiting research period at the Φ-lab starts in March 2020.
Industrial Fellow. Geoville Information Systems and Data Processing GmbH. Winter/Spring 2019.
As an enthusiast on the use of AI for new science and applications, helping to explore and accelerate the integration of disruptive technologies in EO workflows is a huge part of my task in the Φ-lab. Together with the EU Satellite Centre, I am prototyping new Deep Learning and cloud computing based methodologies for infrastructure mapping and monitoring with
Copernicus Sentinel-1 in desert regions. In collaboration with the Italian National Geophysics Institute, I carry-out research in potential ionosphere-lithosphere coupling through a machine learning analysis of Swarm data. To facilitate the acquisition of training data to support machine learning analyses, I am managing the development of a crowdsourcing platform. Finally, I stimulate the assimilation of disruptive technologies, such as quantum computing, AI and virtual reality in EO through the organisation of workshops, trainings app camps and other events.
My background is in mathematics, and have specialised in EO since carrying-out an MSc in remote sensing and image interpretation at Edinburgh University. While working at ESA, I obtained a PhD in geoinformation at the University of Tor Vergata in Rome, focusing on SAR applications.
Bachelor degree in Computer Engineering at the University of Pisa and currently enrolled in the master degree in Computer Engineering Enterprise System of the same university. He decided to focus on machine learning and multimedia data management, but he is also interested in the various new technologies that are continuously emerging in the industry.
Working in a place where people from different countries and cultures cooperate to reach a common goal is illuminating. I am eager to use my skills for the development of tools and algorithms that can be part of a set of instruments used by scientists and third parties to study and understand how our planet evolves, how the climate changes and what we can do to prevent calamities, and also how to position the crop fields to optimise the food production.
I am working in the Φ-lab as an external researcher, and will be carrying out my Master’s thesis taking part in several projects which concern the computer vision, such as crop and vegetation mapping and data analysis of satellite images. The main purpose is to find and develop algorithms that, with the use of artificial intelligence methods, can extract as much information as possible from the available data.
I have a BSc degree in Engineering in Computer Science from University of Rome La Sapienza and I am currently completing my studies with a MSc in Artificial Intelligence and Robotics Engineering from the same University.
I am interested in the use of physics-based AI applied to EO data, in order to better understand the environmental and ecological interactions underlying the Earth system, particularly in light of environmental challenges such as climate change and ecosystem vulnerability. As an Internal Research Fellow at the Φ-lab, my main research focuses on applying AI to emulate physical radiative transfer models (RTMs). I am currently employing this method for Methane retrieval from Sentinel 5P’s (S5P) TROPospheric Monitoring Instrument (TROPOMI), through using AI to learn the physics of costly RTMs within current retrieval schemes. I also work on building internal Φ-lab capacity for deploying AI at the edge, specifically on the Φ-sat-1 experiment to develop a cloud detection algorithm for hyperspectral data on-board a CubeSat, as well as providing support for hyperspectral data processing and parameter retrieval for forestry and agricultural applications. I have a degree in Geography and an MSc in Remote Sensing (MSc) from University College London (UCL), and a PhD from UCL, co-supervised and funded by the European Commission’s Joint Research Centre, on the use of radiative transfer modelling to underpin uncertainty and traceability for EO products.
Deep Learning (DL) has already provided new and powerful solutions for common problems in the field of computer vision and I am confident that it has the potential to help us with traditional problems in Earth Observation too. In the Φ-lab, I work together with my colleagues and external partners on traditional use cases like crop-type mapping or cloud masking. We hope to find better and more efficient solutions by using new Deep Learning methods. At the same time, I investigate how we can capitalise both on the trend of smaller satellites in space and the new powerful AI-on-the-edge hardware to make new satellites smarter and more powerful.
I have a BSc and a MSc in Meteorology from the University of Hamburg (Germany) and, because my passion is computer science, I have been studying programming in my leisure time since high school.
John Yackel has a Ph.D. from the Center for Earth Observation Science in Winnipeg Canada and has been a Professor of Geography at the University of Calgary Canada since 2000. Phi-lab collaborator spring/summer 2019. He uses multifrequency surface and satellite microwave scatterometer and SAR data to invert snow covered Arctic sea ice geophysical parameters for Arctic climate change assessment.
