Why do we regularise in every iteration for imaging inverse problems?

Re-assign community