by Antoine Houdard, Andrès Almansa, Julie Delon
Abstract:
In this work, we revisit the global denoising framework recently introduced by Talebi & Milanfar, with the classical formalism of diagonal estimation. We analyze the asymptotic behavior of its mean-squared error restoration performance when the image size tends to infinity. We introduce precise conditions both on the image and the global filter to ensure and quantify this convergence. We also discuss open issues concerning the most challenging aspect, namely the extension of these results to the non-oracle case.
Reference:
Demystifying the asymptotic behavior of global denoising (Antoine Houdard, Andrès Almansa, Julie Delon), In Journal of Mathematical Imaging and Vision, Springer Verlag, volume 59, 2017.
Bibtex Entry:
@article{Houdard2017,
Abstract = {In this work, we revisit the global denoising framework recently introduced by Talebi {\&} Milanfar, with the classical formalism of diagonal estimation. We analyze the asymptotic behavior of its mean-squared error restoration performance when the image size tends to infinity. We introduce precise conditions both on the image and the global filter to ensure and quantify this convergence. We also discuss open issues concerning the most challenging aspect, namely the extension of these results to the non-oracle case.},
Annote = {Online: 21 March 2017},
Author = {Houdard, Antoine and Almansa, Andr{\`{e}}s and Delon, Julie},
Doi = {10.1007/s10851-017-0716-6},
Journal = {Journal of Mathematical Imaging and Vision},
Keywords = {Asymptotic study,Diagonal Estimation,Global Denoising,Wiener Filtering},
Number = {3},
Pages = {456--480},
Publisher = {Springer Verlag},
Title = {{Demystifying the asymptotic behavior of global denoising}},
Url = {https://hal.archives-ouvertes.fr/hal-01340822/},
Volume = {59},
Year = {2017},
Bdsk-Url-1 = {https://hal.archives-ouvertes.fr/hal-01340822/},
Bdsk-Url-2 = {https://doi.org/10.1007/s10851-017-0716-6}}