Standard
Kernel Hebbian algorithm for single-frame super-resolution. /
Kim, Kwang In; Franz, Matthias O.; Schölkopf, Bernhard.
Statistical Learning in Computer Vision, ECCV 2004 Workshop, Prague, Czech Republic, May 2004. ed. / A. Leonardis; H. Bischof. Tubingen, Germany: Max Planck Institute fur biologische Kybernetik, 2004. p. 135-149.
Research output: Contribution in Book/Report/Proceedings - With ISBN/ISSN › Conference contribution/Paper › peer-review
Harvard
Kim, KI, Franz, MO & Schölkopf, B 2004,
Kernel Hebbian algorithm for single-frame super-resolution. in A Leonardis & H Bischof (eds),
Statistical Learning in Computer Vision, ECCV 2004 Workshop, Prague, Czech Republic, May 2004. Max Planck Institute fur biologische Kybernetik, Tubingen, Germany, pp. 135-149. <
http://www.is.tuebingen.mpg.de/fileadmin/user_upload/files/publications/pdf2645.pdf>
APA
Kim, K. I., Franz, M. O., & Schölkopf, B. (2004).
Kernel Hebbian algorithm for single-frame super-resolution. In A. Leonardis, & H. Bischof (Eds.),
Statistical Learning in Computer Vision, ECCV 2004 Workshop, Prague, Czech Republic, May 2004 (pp. 135-149). Max Planck Institute fur biologische Kybernetik.
http://www.is.tuebingen.mpg.de/fileadmin/user_upload/files/publications/pdf2645.pdf
Vancouver
Kim KI, Franz MO, Schölkopf B.
Kernel Hebbian algorithm for single-frame super-resolution. In Leonardis A, Bischof H, editors, Statistical Learning in Computer Vision, ECCV 2004 Workshop, Prague, Czech Republic, May 2004. Tubingen, Germany: Max Planck Institute fur biologische Kybernetik. 2004. p. 135-149
Author
Bibtex
@inproceedings{deb29ff235af4949ae6f61c67d2f5c64,
title = "Kernel Hebbian algorithm for single-frame super-resolution",
abstract = "This paper presents a method for single-frame image superresolution using an unsupervised learning technique. The required prior knowledge about the high-resolution images is obtained from Kernel Principal Component Analysis (KPCA). The original form of KPCA, however, can be only applied to strongly restricted image classes due to the limited number of training examples that can be processed. We therefore propose a new iterative method for performing KPCA, the Kernel Hebbian Algorithm. By kernelizing the Generalized Hebbian Algorithm, one can iteratively estimate the Kernel Principal Components with only linear order memory complexity. The resulting super-resolution algorithm shows a comparable performance to the existing supervised methods on images containing faces and natural scenes.",
author = "Kim, {Kwang In} and Franz, {Matthias O.} and Bernhard Sch{\"o}lkopf",
year = "2004",
language = "English",
pages = "135--149",
editor = "Leonardis, {A. } and Bischof, {H. }",
booktitle = "Statistical Learning in Computer Vision, ECCV 2004 Workshop, Prague, Czech Republic, May 2004",
publisher = "Max Planck Institute fur biologische Kybernetik",
}
RIS
TY - GEN
T1 - Kernel Hebbian algorithm for single-frame super-resolution
AU - Kim, Kwang In
AU - Franz, Matthias O.
AU - Schölkopf, Bernhard
PY - 2004
Y1 - 2004
N2 - This paper presents a method for single-frame image superresolution using an unsupervised learning technique. The required prior knowledge about the high-resolution images is obtained from Kernel Principal Component Analysis (KPCA). The original form of KPCA, however, can be only applied to strongly restricted image classes due to the limited number of training examples that can be processed. We therefore propose a new iterative method for performing KPCA, the Kernel Hebbian Algorithm. By kernelizing the Generalized Hebbian Algorithm, one can iteratively estimate the Kernel Principal Components with only linear order memory complexity. The resulting super-resolution algorithm shows a comparable performance to the existing supervised methods on images containing faces and natural scenes.
AB - This paper presents a method for single-frame image superresolution using an unsupervised learning technique. The required prior knowledge about the high-resolution images is obtained from Kernel Principal Component Analysis (KPCA). The original form of KPCA, however, can be only applied to strongly restricted image classes due to the limited number of training examples that can be processed. We therefore propose a new iterative method for performing KPCA, the Kernel Hebbian Algorithm. By kernelizing the Generalized Hebbian Algorithm, one can iteratively estimate the Kernel Principal Components with only linear order memory complexity. The resulting super-resolution algorithm shows a comparable performance to the existing supervised methods on images containing faces and natural scenes.
M3 - Conference contribution/Paper
SP - 135
EP - 149
BT - Statistical Learning in Computer Vision, ECCV 2004 Workshop, Prague, Czech Republic, May 2004
A2 - Leonardis, A.
A2 - Bischof, H.
PB - Max Planck Institute fur biologische Kybernetik
CY - Tubingen, Germany
ER -