Super-Resolution Based on Clustered Examples

doi:10.1142/S0218001416550156

機構典藏 > College of Engineering > Graduate Institute & Department of Computer Science and Information Engineering > Journal Article > Item 987654321/111858

Please use this identifier to cite or link to this item: https://tkuir.lib.tku.edu.tw/dspace/handle/987654321/111858

Title:	Super-Resolution Based on Clustered Examples
Authors:	Ching Ting Tu;Hsiau Wen Lin;Hwei-Jen Lin;Yue Shen Li
Keywords:	Super-resolution;locally linear embedding (LLE);K-means++++ clustering;interpolation;learning-based method;example-based method;degradation
Date:	2016-03-18
Issue Date:	2017-10-27 02:10:39 (UTC+8)
Publisher:	World Scientific Publishing
Abstract:	In this paper, we propose an improved version of the neighbor embedding super-resolution (SR) algorithm proposed by Chang et al. [Super-resolution through neighbor embedding, in Proc. 2004 IEEE Computer Society Conf. Computer Vision and Pattern Recognition(CVPR), Vol. 1 (2004), pp. 275–282]. The neighbor embedding SR algorithm requires intensive computational time when finding the K nearest neighbors for the input patch in a huge set of training samples. We tackle this problem by clustering the training sample into a number of clusters, with which we first find for the input patch the nearest cluster center, and then find the K nearest neighbors in the corresponding cluster. In contrast to Chang’s method, which uses Euclidean distance to find the K nearest neighbors of a low-resolution patch, we define a similarity function and use that to find the K most similar neighbors of a low-resolution patch. We then use local linear embedding (LLE) [S. T. Roweis and L. K. Saul, Nonlinear dimensionality reduction by locally linear embedding, Science 290(5500) (2000) 2323–2326] to find optimal coefficients, with which the linear combination of the K most similar neighbors best approaches the input patch. These coefficients are then used to form a linear combination of the K high-frequency patches corresponding to the K respective low-resolution patches (or the K most similar neighbors). The resulting high-frequency patch is then added to the enlarged (or up-sampled) version of the input patch. Experimental results show that the proposed clustering scheme efficiently reduces computational time without significantly affecting the performance.
Relation:	International Journal of Pattern Recognition and Artificial Intelligence 30(6), p.1655015 (15 pages)
DOI:	10.1142/S0218001416550156
Appears in Collections:	[Graduate Institute & Department of Computer Science and Information Engineering] Journal Article

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	171	View/Open
Super-Resolution Based on Clustered Examples.pdf		2719Kb	Adobe PDF	2	View/Open

Loading...