DEEPLIM – AIKO/ESA – Generative Artificial Intelligence for high performing inversions models

2019-12-27 – 2021-02-26

ESA – European Space Agency

AIKO SRL

The scope of the project is the develompment of Genenerative artificial intelligence models (and in particular of Generative Adversial Neural Netwoks, GANs) in support of high performing inversion models. The use of  GANs can help in increasing the datasets used for training of satellite inversion models. Two tests cases will be analysed in this project: 1) Inversion of thermal infrared radiances from origianal (simulated + measured) and GANs augmented datasets to obtain land surface emissivity and land surface Temperature and 2) vTEC map assessment and Gap filling through the use of GANs.