Image Super-Resolution Via Sparse Representation. Yang, J., Wright, J., Huang, T. S., & Ma, Y. IEEE Transactions on Image Processing, 19(11):2861–2873, November, 2010. Conference Name: IEEE Transactions on Image Processingdoi abstract bibtex This paper presents a new approach to single-image superresolution, based upon sparse signal representation. Research on image statistics suggests that image patches can be well-represented as a sparse linear combination of elements from an appropriately chosen over-complete dictionary. Inspired by this observation, we seek a sparse representation for each patch of the low-resolution input, and then use the coefficients of this representation to generate the high-resolution output. Theoretical results from compressed sensing suggest that under mild conditions, the sparse representation can be correctly recovered from the downsampled signals. By jointly training two dictionaries for the low- and high-resolution image patches, we can enforce the similarity of sparse representations between the low-resolution and high-resolution image patch pair with respect to their own dictionaries. Therefore, the sparse representation of a low-resolution image patch can be applied with the high-resolution image patch dictionary to generate a high-resolution image patch. The learned dictionary pair is a more compact representation of the patch pairs, compared to previous approaches, which simply sample a large amount of image patch pairs , reducing the computational cost substantially. The effectiveness of such a sparsity prior is demonstrated for both general image super-resolution (SR) and the special case of face hallucination. In both cases, our algorithm generates high-resolution images that are competitive or even superior in quality to images produced by other similar SR methods. In addition, the local sparse modeling of our approach is naturally robust to noise, and therefore the proposed algorithm can handle SR with noisy inputs in a more unified framework.
@article{yang_image_2010,
title = {Image {Super}-{Resolution} {Via} {Sparse} {Representation}},
volume = {19},
issn = {1941-0042},
doi = {10.1109/TIP.2010.2050625},
abstract = {This paper presents a new approach to single-image superresolution, based upon sparse signal representation. Research on image statistics suggests that image patches can be well-represented as a sparse linear combination of elements from an appropriately chosen over-complete dictionary. Inspired by this observation, we seek a sparse representation for each patch of the low-resolution input, and then use the coefficients of this representation to generate the high-resolution output. Theoretical results from compressed sensing suggest that under mild conditions, the sparse representation can be correctly recovered from the downsampled signals. By jointly training two dictionaries for the low- and high-resolution image patches, we can enforce the similarity of sparse representations between the low-resolution and high-resolution image patch pair with respect to their own dictionaries. Therefore, the sparse representation of a low-resolution image patch can be applied with the high-resolution image patch dictionary to generate a high-resolution image patch. The learned dictionary pair is a more compact representation of the patch pairs, compared to previous approaches, which simply sample a large amount of image patch pairs , reducing the computational cost substantially. The effectiveness of such a sparsity prior is demonstrated for both general image super-resolution (SR) and the special case of face hallucination. In both cases, our algorithm generates high-resolution images that are competitive or even superior in quality to images produced by other similar SR methods. In addition, the local sparse modeling of our approach is naturally robust to noise, and therefore the proposed algorithm can handle SR with noisy inputs in a more unified framework.},
language = {en},
number = {11},
journal = {IEEE Transactions on Image Processing},
author = {Yang, Jianchao and Wright, John and Huang, Thomas S. and Ma, Yi},
month = nov,
year = {2010},
note = {Conference Name: IEEE Transactions on Image Processing},
keywords = {\#Representation{\textgreater}SR, \#Sparse, /unread, Compressed sensing, Computational efficiency, Dictionaries, Face hallucination, Image generation, Image resolution, Noise robustness, Signal representations, Signal resolution, Statistics, Strontium, image super-resolution (SR), nonnegative matrix factorization, sparse coding, sparse representation, ⭐⭐⭐⭐⭐},
pages = {2861--2873},
}
Downloads: 0
{"_id":"GMgTn8rcc9HqqwTrB","bibbaseid":"yang-wright-huang-ma-imagesuperresolutionviasparserepresentation-2010","downloads":0,"creationDate":"2015-09-03T07:18:47.887Z","title":"Image Super-Resolution Via Sparse Representation","author_short":["Yang, J.","Wright, J.","Huang, T. S.","Ma, Y."],"year":2010,"bibtype":"article","biburl":"https://bibbase.org/zotero/zzhenry2012","bibdata":{"bibtype":"article","type":"article","title":"Image Super-Resolution Via Sparse Representation","volume":"19","issn":"1941-0042","doi":"10.1109/TIP.2010.2050625","abstract":"This paper presents a new approach to single-image superresolution, based upon sparse signal representation. Research on image statistics suggests that image patches can be well-represented as a sparse linear combination of elements from an appropriately chosen over-complete dictionary. Inspired by this observation, we seek a sparse representation for each patch of the low-resolution input, and then use the coefficients of this representation to generate the high-resolution output. Theoretical results from compressed sensing suggest that under mild conditions, the sparse representation can be correctly recovered from the downsampled signals. By jointly training two dictionaries for the low- and high-resolution image patches, we can enforce the similarity of sparse representations between the low-resolution and high-resolution image patch pair with respect to their own dictionaries. Therefore, the sparse representation of a low-resolution image patch can be applied with the high-resolution image patch dictionary to generate a high-resolution image patch. The learned dictionary pair is a more compact representation of the patch pairs, compared to previous approaches, which simply sample a large amount of image patch pairs , reducing the computational cost substantially. The effectiveness of such a sparsity prior is demonstrated for both general image super-resolution (SR) and the special case of face hallucination. In both cases, our algorithm generates high-resolution images that are competitive or even superior in quality to images produced by other similar SR methods. In addition, the local sparse modeling of our approach is naturally robust to noise, and therefore the proposed algorithm can handle SR with noisy inputs in a more unified framework.","language":"en","number":"11","journal":"IEEE Transactions on Image Processing","author":[{"propositions":[],"lastnames":["Yang"],"firstnames":["Jianchao"],"suffixes":[]},{"propositions":[],"lastnames":["Wright"],"firstnames":["John"],"suffixes":[]},{"propositions":[],"lastnames":["Huang"],"firstnames":["Thomas","S."],"suffixes":[]},{"propositions":[],"lastnames":["Ma"],"firstnames":["Yi"],"suffixes":[]}],"month":"November","year":"2010","note":"Conference Name: IEEE Transactions on Image Processing","keywords":"#Representation\\textgreaterSR, #Sparse, /unread, Compressed sensing, Computational efficiency, Dictionaries, Face hallucination, Image generation, Image resolution, Noise robustness, Signal representations, Signal resolution, Statistics, Strontium, image super-resolution (SR), nonnegative matrix factorization, sparse coding, sparse representation, ⭐⭐⭐⭐⭐","pages":"2861–2873","bibtex":"@article{yang_image_2010,\n\ttitle = {Image {Super}-{Resolution} {Via} {Sparse} {Representation}},\n\tvolume = {19},\n\tissn = {1941-0042},\n\tdoi = {10.1109/TIP.2010.2050625},\n\tabstract = {This paper presents a new approach to single-image superresolution, based upon sparse signal representation. Research on image statistics suggests that image patches can be well-represented as a sparse linear combination of elements from an appropriately chosen over-complete dictionary. Inspired by this observation, we seek a sparse representation for each patch of the low-resolution input, and then use the coefficients of this representation to generate the high-resolution output. Theoretical results from compressed sensing suggest that under mild conditions, the sparse representation can be correctly recovered from the downsampled signals. By jointly training two dictionaries for the low- and high-resolution image patches, we can enforce the similarity of sparse representations between the low-resolution and high-resolution image patch pair with respect to their own dictionaries. Therefore, the sparse representation of a low-resolution image patch can be applied with the high-resolution image patch dictionary to generate a high-resolution image patch. The learned dictionary pair is a more compact representation of the patch pairs, compared to previous approaches, which simply sample a large amount of image patch pairs , reducing the computational cost substantially. The effectiveness of such a sparsity prior is demonstrated for both general image super-resolution (SR) and the special case of face hallucination. In both cases, our algorithm generates high-resolution images that are competitive or even superior in quality to images produced by other similar SR methods. In addition, the local sparse modeling of our approach is naturally robust to noise, and therefore the proposed algorithm can handle SR with noisy inputs in a more unified framework.},\n\tlanguage = {en},\n\tnumber = {11},\n\tjournal = {IEEE Transactions on Image Processing},\n\tauthor = {Yang, Jianchao and Wright, John and Huang, Thomas S. and Ma, Yi},\n\tmonth = nov,\n\tyear = {2010},\n\tnote = {Conference Name: IEEE Transactions on Image Processing},\n\tkeywords = {\\#Representation{\\textgreater}SR, \\#Sparse, /unread, Compressed sensing, Computational efficiency, Dictionaries, Face hallucination, Image generation, Image resolution, Noise robustness, Signal representations, Signal resolution, Statistics, Strontium, image super-resolution (SR), nonnegative matrix factorization, sparse coding, sparse representation, ⭐⭐⭐⭐⭐},\n\tpages = {2861--2873},\n}\n\n\n\n","author_short":["Yang, J.","Wright, J.","Huang, T. S.","Ma, Y."],"key":"yang_image_2010","id":"yang_image_2010","bibbaseid":"yang-wright-huang-ma-imagesuperresolutionviasparserepresentation-2010","role":"author","urls":{},"keyword":["#Representation\\textgreaterSR","#Sparse","/unread","Compressed sensing","Computational efficiency","Dictionaries","Face hallucination","Image generation","Image resolution","Noise robustness","Signal representations","Signal resolution","Statistics","Strontium","image super-resolution (SR)","nonnegative matrix factorization","sparse coding","sparse representation","⭐⭐⭐⭐⭐"],"metadata":{"authorlinks":{}},"downloads":0,"html":""},"search_terms":["image","super","resolution","via","sparse","representation","yang","wright","huang","ma"],"keywords":["#representation\\textgreatersr","#sparse","/unread","compressed sensing","computational efficiency","dictionaries","face hallucination","image generation","image resolution","noise robustness","signal representations","signal resolution","statistics","strontium","image super-resolution (sr)","nonnegative matrix factorization","sparse coding","sparse representation","⭐⭐⭐⭐⭐"],"authorIDs":[],"dataSources":["9cexBw6hrwgyZphZZ","nZHrFJKyxKKDaWYM8"]}