TY - GEN

T1 - Efficient learning-based image enhancement

T2 - application to super-resolution and compression artifact removal

AU - Kwon, Younghee

AU - Kim, Kwang In

AU - Kim, Jin H.

AU - Theobalt, Christian

PY - 2012

Y1 - 2012

N2 - In this paper, we describe a framework for learning-based image enhancement. At the core of our algorithm lies a generic regularization framework that comprises a prior on natural images, as well as an application-specific conditional model based on Gaussian processes. In contrast to prior learning-based approaches, our algorithm can instantly learn task-specific degradation models from sample images which enables users to easily adopt the algorithm to a specific problem and data set of interest. This is facilitated by ourefficient approximation scheme of large-scale Gaussian processes. We demonstrate the efficiency and effectiveness of our approach by applying it to two example enhancement applications: single-image super-resolution as well as artifact removal in JPEG-encoded images.

AB - In this paper, we describe a framework for learning-based image enhancement. At the core of our algorithm lies a generic regularization framework that comprises a prior on natural images, as well as an application-specific conditional model based on Gaussian processes. In contrast to prior learning-based approaches, our algorithm can instantly learn task-specific degradation models from sample images which enables users to easily adopt the algorithm to a specific problem and data set of interest. This is facilitated by ourefficient approximation scheme of large-scale Gaussian processes. We demonstrate the efficiency and effectiveness of our approach by applying it to two example enhancement applications: single-image super-resolution as well as artifact removal in JPEG-encoded images.

U2 - 10.5244/C.26.14

DO - 10.5244/C.26.14

M3 - Conference contribution/Paper

SP - 14.1-14.12

BT - Proc. British Machine Vision Conference (BMVC) 2012

ER -