Environmental Computational Science and Earth Observation Laboratory
Machine learning to understand the environment.
We are surrounded by sensors acquiring data about our environment. From the mobile phone in our pocket, the CCTV cameras, up to flying sensors like drones or satellites, the potential of monitoring and understanding the environment is untapped.
But these data are massive and acquired at a rapid pace, and no single person can make sense of it, especially when thinking globally. Therefore, it becomes necessary to develop technology to organize, catalog, search and process this data. At the Environmental Computational Science and Earth Observation Laboratory (ECEO), we extract knowledge from data that are heterogeneous and often very unstructured, and acquired at multiple scales by a number of imaging devices. We do so with machine learning algorithms – from classical feature extraction to complex deep neural networks – and develop new algorithms to make sense of our Earth.
We then use our algorithms to innovate in environmental science: from animal conservation, to urban studies, from snow monitoring to land change detection. With Artificial intelligence and Earth Observation, ECEO designs new ways to see, understand and impact our blue planet.
