Abstract : Plug-and-play (PnP) methods are a class of iterative algorithms for imaging inverse problems, leveraging off-the-shelf Gaussian denoisers for regularization. These methods achieve impressive visual results, especially when deep neural networks parameterize the denoisers. However, the theoretical convergence of PnP methods has yet to be fully established. This talk provides an overview of the PnP literature, introduces new convergence results for PnP algorithms when paired with specific denoisers, and finally, presents a novel Bregman version of Plug-and-Play.
Idriss Mazari
