A 36-month, fully funded PhD position is available at CREATIS, starting September 2025.
Objectives In this thesis, we will focus on the design of i) the direct model and ii) the reconstruction methods for which the direct model is not known in advance. This problem is related to the choice of the projection matrix in compressed sensing. While compressed sensing is non adaptive, the adaptive design of the direct model has remained largely unexplored, in particular when complementary acquisitions are available.
Keywords Image reconstruction, optimization, computational optics, deep learning.
Offer details PDF