Skip to main content

A spectral based visual matching method for image classification

Song, Yan and Guo, Wu and Dai, Li-Rong and McLoughlin, Ian Vince (2014) A spectral based visual matching method for image classification. In: 2014 International Conference on Audio, Language and Image Processing. IEEE, pp. 666-670. ISBN 978-1-4799-3902-2. E-ISBN 978-1-4799-3903-9. (doi:10.1109/ICALIP.2014.7009878) (The full text of this publication is not currently available from this repository. You may be able to access a copy if URLs are provided) (KAR id:48921)

The full text of this publication is not currently available from this repository. You may be able to access a copy if URLs are provided.
Official URL:


Visual matching algorithms can be described in terms of visual content representation and similarity measure. With local feature based representations, visual matching can be restated as: (1) how to obtain visual similarity from the local kernel matrix, and (2) how to calculate the local kernel matrix effectively and efficiently. Existing methods mostly focus on the former, and use Euclidean distance to calculate the local kernel under Gaussian noise assumption. However, this assumption may not be optimal for gradient based local features. In this paper, we propose a Local Coding based Spectral Analysis (LCSA) method to exploit the low dimensional manifold structure in feature space. Specifically, we select a set of anchor points, and represent each feature as a linear combination of anchor points with locality constraint. The spectral analysis can then be efficiently processed according to this representation. Following the derivation of Efficient Match Kernel (EMK) [6], a compact lower-dimensional set-level image representation is obtained for visual similarity measure. Experimental results on several benchmark image classification datasets, i.e. 15-scenes and Caltech101/256, show superior performance compared with the existing state-of-the-art techniques with SIFT feature.

Item Type: Book section
DOI/Identification number: 10.1109/ICALIP.2014.7009878
Uncontrolled keywords: Bag-of-Visual Words, EMK derivation, Encoding, Euclidean distance, Feature extraction, Gaussian noise, Graph Embedding, Image classification, Kernel, LCSA method, Match Kernel, SIFT feature space, Spectral analysis, Vectors, Visualization, anchor point linear combination, compact lower-dimensional set-level image represen, efficient match kernel derivation, feature based representation, feature extraction, image classification, image coding, image matching, image representation, local coding based spectral analysis method, local kernel matrix, matrix algebra, spectral analysis, spectral based visual matching method, visual content representation, visual matching
Subjects: T Technology
Divisions: Divisions > Division of Computing, Engineering and Mathematical Sciences > School of Computing
Depositing User: Ian McLoughlin
Date Deposited: 25 Aug 2015 08:50 UTC
Last Modified: 17 Aug 2022 10:58 UTC
Resource URI: (The current URI for this page, for reference purposes)
McLoughlin, Ian Vince:
  • Depositors only (login required):