%0 Journal Article %1 Polatkan2015 %A Polatkan, Gungor %A Zhou, Mingyuan %A Carin, Lawrence %A Blei, David %A Daubechies, Ingrid %D 2015 %J IEEE Transactions on Pattern Analysis and Machine Intelligence %K analysis bayesian microscopy software superresolution %N 2 %P 346--358 %R 10.1109/TPAMI.2014.2321404 %T A bayesian nonparametric approach to image super-resolution %U https://doi.org/10.1101/2020.02.15.950873 %V 37 %X Super-resolution methods form high-resolution images from low-resolution images. In this paper, we develop a new Bayesian nonparametric model for super-resolution. Our method uses a beta-Bernoulli process to learn a set of recurring visual patterns, called dictionary elements, from the data. Because it is nonparametric, the number of elements found is also determined from the data. We test the results on both benchmark and natural images, comparing with several other models from the research literature. We perform large-scale human evaluation experiments to assess the visual quality of the results. In a first implementation, we use Gibbs sampling to approximate the posterior. However, this algorithm is not feasible for large-scale data. To circumvent this, we then develop an online variational Bayes (VB) algorithm. This algorithm finds high quality dictionaries in a fraction of the time needed by the Gibbs sampler.

@article{Polatkan2015, abstract = {Super-resolution methods form high-resolution images from low-resolution images. In this paper, we develop a new Bayesian nonparametric model for super-resolution. Our method uses a beta-Bernoulli process to learn a set of recurring visual patterns, called dictionary elements, from the data. Because it is nonparametric, the number of elements found is also determined from the data. We test the results on both benchmark and natural images, comparing with several other models from the research literature. We perform large-scale human evaluation experiments to assess the visual quality of the results. In a first implementation, we use Gibbs sampling to approximate the posterior. However, this algorithm is not feasible for large-scale data. To circumvent this, we then develop an online variational Bayes (VB) algorithm. This algorithm finds high quality dictionaries in a fraction of the time needed by the Gibbs sampler.}, added-at = {2020-03-27T10:41:29.000+0100}, archiveprefix = {arXiv}, arxivid = {1209.5019}, author = {Polatkan, Gungor and Zhou, Mingyuan and Carin, Lawrence and Blei, David and Daubechies, Ingrid}, biburl = {https://www.bibsonomy.org/bibtex/21e4c21229a3b61e80ef53af4ec9f9c5b/kfriedl}, doi = {10.1109/TPAMI.2014.2321404}, eprint = {1209.5019}, file = {:C$\backslash$:/Users/Karoline/AppData/Local/Mendeley Ltd./Mendeley Desktop/Downloaded/Gabitto et al. - Unknown - A BAYESIAN NONPARAMETRIC APPROACH TO SUPER-RESOLUTION SINGLE-MOLECULE LOCALIZATION.pdf:pdf}, interhash = {8fb02a15d0343a974ad1dddc77430b3a}, intrahash = {1e4c21229a3b61e80ef53af4ec9f9c5b}, issn = {01628828}, journal = {IEEE Transactions on Pattern Analysis and Machine Intelligence}, keywords = {analysis bayesian microscopy software superresolution}, number = 2, pages = {346--358}, timestamp = {2020-04-06T15:12:33.000+0200}, title = {{A bayesian nonparametric approach to image super-resolution}}, url = {https://doi.org/10.1101/2020.02.15.950873}, volume = 37, year = 2015 }