• In general, the dictionary learning based image super-resolution methods adopts the patch based reconstruction of the target image. Such algorithms are very slow as it solves the convex optimization problem individually for each patch in the image. For example, sparse representation based image super-resolution methods are good in terms of output whilst they require high-time to perform.
    • In such situation, is there a method which will help to use sparse representation method for image super-resolution but it will reduce the time required by adopting the method.

    Similar questions and discussions