Structured and Localized Image Restoration
This work addresses image restoration for computer vision applications, presenting a method with strong statistical guarantees but incremental improvements over existing techniques.
The authors tackled image restoration by developing a novel approach that combines localized structured prediction with non-linear multi-task learning, optimizing a penalized energy function regularized by patch distances to an external clean database. They demonstrated practical effectiveness on standard benchmarks for various image restoration problems.
We present a novel approach to image restoration that leverages ideas from localized structured prediction and non-linear multi-task learning. We optimize a penalized energy function regularized by a sum of terms measuring the distance between patches to be restored and clean patches from an external database gathered beforehand. The resulting estimator comes with strong statistical guarantees leveraging local dependency properties of overlapping patches. We derive the corresponding algorithms for energies based on the mean-squared and Euclidean norm errors. Finally, we demonstrate the practical effectiveness of our model on different image restoration problems using standard benchmarks.