IRD opens a state paid permanent position of research associate

IRD, the French Research Institute for Development, opens a state paid permanent position of research associate in "Big data and environmental analytics" ( The AMAP team working on tropical forest remote sensing is one of the possible host labs (, which is seeking a specialist in signal analysis, familiar with applications in remote sensing of vegetation, especially based on learning methods to work closely with modelers and ecologists who develop models of structure and function of tropical forest vegetation. More precisely, the constant increase in the volume, revisitability, diversity and dimensionality of remote sensing data necessitates the development of methodological innovations that must not only allow effective multimodal data processing, but be able to adapt to variations in data distributions. These can result from differences in satellite data acquisition conditions (atmospheric characteristics, illumination), on-board sensor specificities (number of bands, spatial resolution), shooting angles, and so on. This problem can be addressed in the theoretical framework of domain adaptation and more generally transfer learning. The complementarity of the data and the multimodal data fusion combining multi-sensor and multidate data is particularly relevant to characterize the multi-scale, spatial and temporal dynamics of tropical forests (seasonal phenology, forest turnover and carbon fluxes) that appear progressively as keys or even determinants of the functional biological diversity of these biomes (relation between dynamics, composition and functioning). The radiative transfer simulation in 3D scenes reconstructed from LiDAR data facilitates the inversion of the signal and allows training models that should in the near future enable the large-scale functional mapping of forest canopies via terrestrial, airborne (including drone) and / or satellite sensors.

Pour en savoir + :

 Retour aux actualités