AstroDeep

AstroDeep gathers a team of researchers applying Bayesian deep nets to cosmology. This ANR  project is lead by APC, Paris and includes members from CosmoStat, UPenn and Loria. CosmoStat's participation is mainly focused on using deep learning to measure galaxy shapes and identify blended sources in survey data.

Find out more about this project on the AstroDeep website.