Abstract

In this talk, we will present our ongoing activities in learning better models for inverse problems in imaging. We consider classical variational models used for inverse problems but generalise these models by introducing a large number of free model parameters. We learn the free model parameters by minimising a loss function comparing the reconstructed images obtained from the variational models with ground truth solutions from a training data base. We will also show recent results on learning “deeper” regularisers that are allowed to change their parameters in each iteration of the algorithm. We show applications to different inverse problems in imaging, where we put a particular focus on joint image demosaicing and denoising.