by Cecilia Aguerrebere, Andres Almansa, Julie Delon, Yann Gousseau, Pablo Muse
Abstract:
Recently, impressive denoising results have been achieved by Bayesian approaches which assume Gaussian models for the image patches. This improvement in performance can be attributed to the use of per-patch models. Unfortunately such an approach is particularly unstable for most inverse problems beyond denoising. In this work, we propose the use of a hyperprior to model image patches, in order to stabilize the estimation procedure. There are two main advantages to the proposed restoration scheme: Firstly it is adapted to diagonal degradation matrices, and in particular to missing data problems (e.g. inpainting of missing pixels or zooming). Secondly it can deal with signal dependent noise models, particularly suited to digital cameras. As such, the scheme is especially adapted to computational photography. In order to illustrate this point, we provide an application to high dynamic range imaging from a single image taken with a modified sensor, which shows the effectiveness of the proposed scheme.
Reference:
A Bayesian Hyperprior Approach for Joint Image Denoising and Interpolation, With an Application to HDR Imaging (Cecilia Aguerrebere, Andres Almansa, Julie Delon, Yann Gousseau, Pablo Muse), In IEEE Transactions on Computational Imaging, volume 3, 2017.
Bibtex Entry:
@article{Aguerrebere2014b,
Abstract = {Recently, impressive denoising results have been achieved by Bayesian approaches which assume Gaussian models for the image patches. This improvement in performance can be attributed to the use of per-patch models. Unfortunately such an approach is particularly unstable for most inverse problems beyond denoising. In this work, we propose the use of a hyperprior to model image patches, in order to stabilize the estimation procedure. There are two main advantages to the proposed restoration scheme: Firstly it is adapted to diagonal degradation matrices, and in particular to missing data problems (e.g. inpainting of missing pixels or zooming). Secondly it can deal with signal dependent noise models, particularly suited to digital cameras. As such, the scheme is especially adapted to computational photography. In order to illustrate this point, we provide an application to high dynamic range imaging from a single image taken with a modified sensor, which shows the effectiveness of the proposed scheme.},
Annote = {https://hal.archives-ouvertes.fr/hal-01107519},
Archiveprefix = {arXiv},
Arxivid = {1706.03261},
Author = {Aguerrebere, Cecilia and Almansa, Andres and Delon, Julie and Gousseau, Yann and Muse, Pablo},
Date-Modified = {2019-09-06 16:15:13 +0200},
Doi = {10.1109/TCI.2017.2704439},
Eprint = {1706.03261},
Issn = {2333-9403},
Journal = {IEEE Transactions on Computational Imaging},
Keywords = {Bayesian restoration,Gaussian Mixture Models,Maximum a Posteriori,Non-local patch-based restoration,conjugate distributions,hyper-prior},
Month = {dec},
Number = {4},
Pages = {633--646},
Title = {{A Bayesian Hyperprior Approach for Joint Image Denoising and Interpolation, With an Application to HDR Imaging}},
Url = {https://nounsse.github.io/HBE_project/},
Volume = {3},
Year = {2017},
Bdsk-Url-1 = {https://nounsse.github.io/HBE%7B%5C_%7Dproject/},
Bdsk-Url-2 = {https://doi.org/10.1109/TCI.2017.2704439}}