We will present the theory of functional lifting for general problems with convex regularizers, in the scalar setting. Then, we will discuss the possibilities of extending the framework to vector-valued problems, depending on their structure. In some cases we can show that it leads to computationally heavy but still tractable formulations. These are joint works with D. Cremers, T. Pock, E. Strekalovskiy.
Back to Convex Relaxation Methods for Geometric Problems in Scientific Computing