Abstract
The programme aims at training professionals trained in the use of geospatial data coming from satellite for Earth Observation applications, such as the monitoring of the UN Sustainable Development Goals. It is based on a virtual mobility that provides the basic of machine learning techniques applied to satellite data sets and challenges the participants to provide answers to data processing questions in groups of students by different institutions.
Introduction
The programme aims at training professionals trained in the use of geospatial data coming from satellite for Earth Observation applications, such as the monitoring of the UN Sustainable Development Goals. It is based on a virtual mobility that provides the basic of machine learning techniques applied to satellite data sets and challenges the participants to provide answers to data processing questions in groups of students by different institutions collaborating on-line.
The challenges, to be completed on an open cloud computing platform, refer to environmental monitoring problems in a real urban environment, and will be selected in accordance to a more than decadal experience by some of the instructors in the organizing group in the organization of international contests using EO data sets.
By learning the basic of EO data processing, experiencing the limits and advantages of machine learnings techniques applied to satellite imagery, and solving real problems for a real environment, the virtual component of this blended programme is meant to provide the participants with new abilities, as well as a first example of both distance learning, which will prove important for their future professional career, and team working in a remote environment, which is becoming more and more widespread option for technical research and development, long before COVID.
The in-person part, organized in the urban area of Cremona, will complete the training by learning and the training by on- line doing with some practical activities on site, such as ground truth collection and result validation. This last part will therefore complete the training on the cycle of problem individuation, scenario simulation and output data analysis, and finally verification of the selected solution and the corresponding recorded data.
Partners
- University of Pavia (Paolo Gamba) - Coordinator
- University of Iceland (Gabriele Cavallaro)
- Institut Polytechnique de Grenoble (Jocelyn Chanussot)
- Technische Universität Berlin (Demir Begum)
Material
.jpeg)
