Bayesian K-SVD Using Fast Variational Inference. Serra, J. G., Testa, M., Molina, R., & Katsaggelos, A. K. IEEE Transactions on Image Processing, 26(7):3344–3359, jul, 2017. Paper doi abstract bibtex Recent work in signal processing in general and image processing in particular deals with sparse representation related problems. Two such problems are of paramount importance: an overriding need for designing a well-suited overcomplete dictionary containing a redundant set of atoms-i.e., basis signals-and how to find a sparse representation of a given signal with respect to the chosen dictionary. Dictionary learning techniques, among which we find the popular K-singular value decomposition algorithm, tackle these problems by adapting a dictionary to a set of training data. A common drawback of such techniques is the need for parameter-tuning. In order to overcome this limitation, we propose a fullyautomated Bayesian method that considers the uncertainty of the estimates and produces a sparse representation of the data without prior information on the number of non-zeros in each representation vector. We follow a Bayesian approach that uses a three-tiered hierarchical prior to enforce sparsity on the representations and develop an efficient variational inference framework that reduces computational complexity. Furthermore, we describe a greedy approach that speeds up the whole process. Finally, we present experimental results that show superior performance on two different applications with real images: denoising and inpainting.
@article{Juan2017a,
abstract = {Recent work in signal processing in general and image processing in particular deals with sparse representation related problems. Two such problems are of paramount importance: an overriding need for designing a well-suited overcomplete dictionary containing a redundant set of atoms-i.e., basis signals-and how to find a sparse representation of a given signal with respect to the chosen dictionary. Dictionary learning techniques, among which we find the popular K-singular value decomposition algorithm, tackle these problems by adapting a dictionary to a set of training data. A common drawback of such techniques is the need for parameter-tuning. In order to overcome this limitation, we propose a fullyautomated Bayesian method that considers the uncertainty of the estimates and produces a sparse representation of the data without prior information on the number of non-zeros in each representation vector. We follow a Bayesian approach that uses a three-tiered hierarchical prior to enforce sparsity on the representations and develop an efficient variational inference framework that reduces computational complexity. Furthermore, we describe a greedy approach that speeds up the whole process. Finally, we present experimental results that show superior performance on two different applications with real images: denoising and inpainting.},
author = {Serra, Juan G. and Testa, Matteo and Molina, Rafael and Katsaggelos, Aggelos K.},
doi = {10.1109/TIP.2017.2681436},
issn = {1057-7149},
journal = {IEEE Transactions on Image Processing},
keywords = {Bayesian modeling,Denoising,Dictionary learning,Inpainting,Sparse representation,Variational inference,k-SVD},
month = {jul},
number = {7},
pages = {3344--3359},
pmid = {28362587},
title = {{Bayesian K-SVD Using Fast Variational Inference}},
url = {http://ieeexplore.ieee.org/document/7875464/},
volume = {26},
year = {2017}
}
Downloads: 0
{"_id":"YsRk6znNPdZaaJxM5","bibbaseid":"serra-testa-molina-katsaggelos-bayesianksvdusingfastvariationalinference-2017","author_short":["Serra, J. G.","Testa, M.","Molina, R.","Katsaggelos, A. K."],"bibdata":{"bibtype":"article","type":"article","abstract":"Recent work in signal processing in general and image processing in particular deals with sparse representation related problems. Two such problems are of paramount importance: an overriding need for designing a well-suited overcomplete dictionary containing a redundant set of atoms-i.e., basis signals-and how to find a sparse representation of a given signal with respect to the chosen dictionary. Dictionary learning techniques, among which we find the popular K-singular value decomposition algorithm, tackle these problems by adapting a dictionary to a set of training data. A common drawback of such techniques is the need for parameter-tuning. In order to overcome this limitation, we propose a fullyautomated Bayesian method that considers the uncertainty of the estimates and produces a sparse representation of the data without prior information on the number of non-zeros in each representation vector. We follow a Bayesian approach that uses a three-tiered hierarchical prior to enforce sparsity on the representations and develop an efficient variational inference framework that reduces computational complexity. Furthermore, we describe a greedy approach that speeds up the whole process. Finally, we present experimental results that show superior performance on two different applications with real images: denoising and inpainting.","author":[{"propositions":[],"lastnames":["Serra"],"firstnames":["Juan","G."],"suffixes":[]},{"propositions":[],"lastnames":["Testa"],"firstnames":["Matteo"],"suffixes":[]},{"propositions":[],"lastnames":["Molina"],"firstnames":["Rafael"],"suffixes":[]},{"propositions":[],"lastnames":["Katsaggelos"],"firstnames":["Aggelos","K."],"suffixes":[]}],"doi":"10.1109/TIP.2017.2681436","issn":"1057-7149","journal":"IEEE Transactions on Image Processing","keywords":"Bayesian modeling,Denoising,Dictionary learning,Inpainting,Sparse representation,Variational inference,k-SVD","month":"jul","number":"7","pages":"3344–3359","pmid":"28362587","title":"Bayesian K-SVD Using Fast Variational Inference","url":"http://ieeexplore.ieee.org/document/7875464/","volume":"26","year":"2017","bibtex":"@article{Juan2017a,\nabstract = {Recent work in signal processing in general and image processing in particular deals with sparse representation related problems. Two such problems are of paramount importance: an overriding need for designing a well-suited overcomplete dictionary containing a redundant set of atoms-i.e., basis signals-and how to find a sparse representation of a given signal with respect to the chosen dictionary. Dictionary learning techniques, among which we find the popular K-singular value decomposition algorithm, tackle these problems by adapting a dictionary to a set of training data. A common drawback of such techniques is the need for parameter-tuning. In order to overcome this limitation, we propose a fullyautomated Bayesian method that considers the uncertainty of the estimates and produces a sparse representation of the data without prior information on the number of non-zeros in each representation vector. We follow a Bayesian approach that uses a three-tiered hierarchical prior to enforce sparsity on the representations and develop an efficient variational inference framework that reduces computational complexity. Furthermore, we describe a greedy approach that speeds up the whole process. Finally, we present experimental results that show superior performance on two different applications with real images: denoising and inpainting.},\nauthor = {Serra, Juan G. and Testa, Matteo and Molina, Rafael and Katsaggelos, Aggelos K.},\ndoi = {10.1109/TIP.2017.2681436},\nissn = {1057-7149},\njournal = {IEEE Transactions on Image Processing},\nkeywords = {Bayesian modeling,Denoising,Dictionary learning,Inpainting,Sparse representation,Variational inference,k-SVD},\nmonth = {jul},\nnumber = {7},\npages = {3344--3359},\npmid = {28362587},\ntitle = {{Bayesian K-SVD Using Fast Variational Inference}},\nurl = {http://ieeexplore.ieee.org/document/7875464/},\nvolume = {26},\nyear = {2017}\n}\n","author_short":["Serra, J. G.","Testa, M.","Molina, R.","Katsaggelos, A. K."],"key":"Juan2017a","id":"Juan2017a","bibbaseid":"serra-testa-molina-katsaggelos-bayesianksvdusingfastvariationalinference-2017","role":"author","urls":{"Paper":"http://ieeexplore.ieee.org/document/7875464/"},"keyword":["Bayesian modeling","Denoising","Dictionary learning","Inpainting","Sparse representation","Variational inference","k-SVD"],"metadata":{"authorlinks":{}}},"bibtype":"article","biburl":"https://sites.northwestern.edu/ivpl/files/2023/06/IVPL_Updated_publications-1.bib","dataSources":["KTWAakbPXLGfYseXn","ePKPjG8C6yvpk4mEK","ya2CyA73rpZseyrZ8","D8k2SxfC5dKNRFgro","7Dwzbxq93HWrJEhT6","qhF8zxmGcJfvtdeAg","fvDEHD49E2ZRwE3fb","H7crv8NWhZup4d4by","DHqokWsryttGh7pJE","vRJd4wNg9HpoZSMHD","sYxQ6pxFgA59JRhxi","w2WahSbYrbcCKBDsC","XasdXLL99y5rygCmq","3gkSihZQRfAD2KBo3","t5XMbyZbtPBo4wBGS","bEpHM2CtrwW2qE8FP","teJzFLHexaz5AQW5z"],"keywords":["bayesian modeling","denoising","dictionary learning","inpainting","sparse representation","variational inference","k-svd"],"search_terms":["bayesian","svd","using","fast","variational","inference","serra","testa","molina","katsaggelos"],"title":"Bayesian K-SVD Using Fast Variational Inference","year":2017}