Abstract

Super-resolution algorithms are used to improve the qualityand resolution of low-resolution images. These algorithms can be dividedinto two classes of hallucination- and reconstruction-based ones. Theimprovement factors of these algorithms are limited, however, previousresearch [10], [9] has shown that combining super-resolution algorithmsfrom these two different groups can push the improvement factor further.We have shown in this paper that combining super-resolution algorithmsof the same class can also push the improvement factor up. For thispurpose, we have combined two hallucination based algorithms, namelythe one found in Single Image Super-Resolution from Transformed Self-Exemplars [7] and the Super-Resolution Convolutional Neural Networkfrom [4]. The combination of these two, through an alpha-blending, hasresulted in a system that outperforms state-of-the-art super-resolutionalgorithms on public benchmark datasets.
Original languageEnglish
Title of host publicationVideo Analytics : Face and Facial Expression Recognition and Audience Measurement
PublisherSpringer
Publication date28 Mar 2017
ISBN (Print)978-3-319-56686-3
ISBN (Electronic)978-3-319-56687-0
DOIs
Publication statusPublished - 28 Mar 2017
Event23rd international conference on pattern recognition (ICPR 2016) - Cancun, Mexico
Duration: 4 Dec 20168 Dec 2016

Conference

Conference23rd international conference on pattern recognition (ICPR 2016)
CountryMexico
CityCancun
Period04/12/201608/12/2016
SeriesLecture Notes in Computer Science
Volume10165
ISSN0302-9743

Fingerprint

Optical resolving power
Image resolution

Cite this

Aakerberg, A., Rasmussen, C. B., Nasrollahi, K., & Moeslund, T. B. (2017). Complementing SRCNN by Transformed Self-Exemplars. In Video Analytics: Face and Facial Expression Recognition and Audience Measurement Springer. Lecture Notes in Computer Science, Vol.. 10165 https://doi.org/10.1007/978-3-319-56687-0_11
Aakerberg, Andreas ; Rasmussen, Christoffer Bøgelund ; Nasrollahi, Kamal ; Moeslund, Thomas B. / Complementing SRCNN by Transformed Self-Exemplars. Video Analytics: Face and Facial Expression Recognition and Audience Measurement. Springer, 2017. (Lecture Notes in Computer Science, Vol. 10165).
@inproceedings{cd593f5504804669a755a03962086e89,
title = "Complementing SRCNN by Transformed Self-Exemplars",
abstract = "Super-resolution algorithms are used to improve the qualityand resolution of low-resolution images. These algorithms can be dividedinto two classes of hallucination- and reconstruction-based ones. Theimprovement factors of these algorithms are limited, however, previousresearch [10], [9] has shown that combining super-resolution algorithmsfrom these two different groups can push the improvement factor further.We have shown in this paper that combining super-resolution algorithmsof the same class can also push the improvement factor up. For thispurpose, we have combined two hallucination based algorithms, namelythe one found in Single Image Super-Resolution from Transformed Self-Exemplars [7] and the Super-Resolution Convolutional Neural Networkfrom [4]. The combination of these two, through an alpha-blending, hasresulted in a system that outperforms state-of-the-art super-resolutionalgorithms on public benchmark datasets.",
author = "Andreas Aakerberg and Rasmussen, {Christoffer B{\o}gelund} and Kamal Nasrollahi and Moeslund, {Thomas B.}",
year = "2017",
month = "3",
day = "28",
doi = "10.1007/978-3-319-56687-0_11",
language = "English",
isbn = "978-3-319-56686-3",
series = "Lecture Notes in Computer Science",
publisher = "Springer",
booktitle = "Video Analytics",
address = "Germany",

}

Aakerberg, A, Rasmussen, CB, Nasrollahi, K & Moeslund, TB 2017, Complementing SRCNN by Transformed Self-Exemplars. in Video Analytics: Face and Facial Expression Recognition and Audience Measurement. Springer, Lecture Notes in Computer Science, vol. 10165, 23rd international conference on pattern recognition (ICPR 2016), Cancun, Mexico, 04/12/2016. https://doi.org/10.1007/978-3-319-56687-0_11

Complementing SRCNN by Transformed Self-Exemplars. / Aakerberg, Andreas; Rasmussen, Christoffer Bøgelund; Nasrollahi, Kamal; Moeslund, Thomas B.

Video Analytics: Face and Facial Expression Recognition and Audience Measurement. Springer, 2017. (Lecture Notes in Computer Science, Vol. 10165).

Research output: Contribution to book/anthology/report/conference proceedingArticle in proceedingResearchpeer-review

TY - GEN

T1 - Complementing SRCNN by Transformed Self-Exemplars

AU - Aakerberg, Andreas

AU - Rasmussen, Christoffer Bøgelund

AU - Nasrollahi, Kamal

AU - Moeslund, Thomas B.

PY - 2017/3/28

Y1 - 2017/3/28

N2 - Super-resolution algorithms are used to improve the qualityand resolution of low-resolution images. These algorithms can be dividedinto two classes of hallucination- and reconstruction-based ones. Theimprovement factors of these algorithms are limited, however, previousresearch [10], [9] has shown that combining super-resolution algorithmsfrom these two different groups can push the improvement factor further.We have shown in this paper that combining super-resolution algorithmsof the same class can also push the improvement factor up. For thispurpose, we have combined two hallucination based algorithms, namelythe one found in Single Image Super-Resolution from Transformed Self-Exemplars [7] and the Super-Resolution Convolutional Neural Networkfrom [4]. The combination of these two, through an alpha-blending, hasresulted in a system that outperforms state-of-the-art super-resolutionalgorithms on public benchmark datasets.

AB - Super-resolution algorithms are used to improve the qualityand resolution of low-resolution images. These algorithms can be dividedinto two classes of hallucination- and reconstruction-based ones. Theimprovement factors of these algorithms are limited, however, previousresearch [10], [9] has shown that combining super-resolution algorithmsfrom these two different groups can push the improvement factor further.We have shown in this paper that combining super-resolution algorithmsof the same class can also push the improvement factor up. For thispurpose, we have combined two hallucination based algorithms, namelythe one found in Single Image Super-Resolution from Transformed Self-Exemplars [7] and the Super-Resolution Convolutional Neural Networkfrom [4]. The combination of these two, through an alpha-blending, hasresulted in a system that outperforms state-of-the-art super-resolutionalgorithms on public benchmark datasets.

U2 - 10.1007/978-3-319-56687-0_11

DO - 10.1007/978-3-319-56687-0_11

M3 - Article in proceeding

SN - 978-3-319-56686-3

T3 - Lecture Notes in Computer Science

BT - Video Analytics

PB - Springer

ER -

Aakerberg A, Rasmussen CB, Nasrollahi K, Moeslund TB. Complementing SRCNN by Transformed Self-Exemplars. In Video Analytics: Face and Facial Expression Recognition and Audience Measurement. Springer. 2017. (Lecture Notes in Computer Science, Vol. 10165). https://doi.org/10.1007/978-3-319-56687-0_11