Academic
Publications
Semi-Supervised Dimensionality Reduction Using Pairwise Equivalence Constraints

Semi-Supervised Dimensionality Reduction Using Pairwise Equivalence Constraints,Hakan Cevikalp,Jakob J. Verbeek,Frédéric Jurie,Alexander Kläser

Semi-Supervised Dimensionality Reduction Using Pairwise Equivalence Constraints   (Citations: 9)
BibTex | RIS | RefWorks Download
To deal with the problem of insufficient labeled data, usually side information - given in the form of pairwise equivalence constraints between points - is used to discover groups within data. However, existing methods using side information typically fail in cases with high-dimensional spaces. In this paper, we address the prob- lem of learning from side information for high-dimensional data. To this end, we propose a semi-supervised dimensionality reduction scheme that incorporates pairwise equivalence constraints for finding a better embed- ding space, which improves the performance of subsequent clustering and classification phases. Our method builds on the assumption that points in a sufficiently small neighborhood tend to have the same label. Equiv- alence constraints are employed to modify the neighborhoods and to increase the separability of different classes. Experimental results on high-dimensional image data sets show that integrating side information into the dimensionality reduction improves the clustering and classification performance.
Conference: Computer Vision Theory and Applications - VISAPP , pp. 489-496, 2008
Cumulative Annual
View Publication
The following links allow you to view full publications. These links are maintained by other sources not affiliated with Microsoft Academic Search.
    • ...For instance, faces extracted in successive video frames in roughly same location can be assumed as same person, whereas faces extracted in the same frame in different locations can be seen as different persons [2]...
    • ...Cevikalp et al. [2] proposed pairwise constrained locality preserv� ing projections (CLPP)...

    GX Yuet al. Mixture graph based semi-supervised dimensionality reduction

    • ...To this end, [10] and [3] proposed semisupervised dimensionality reduction methods that exploit both constraints and unconstrained data...

    Su Yanet al. DSP: Robust Semi-supervised Dimensionality Reduction Using Dual Subspa...

    • ...Previously relevant methods [5] [6] [7] [8] implicitly assume that a class consists of a single cluster...
    • ...CLPP [7] builds an affinity matrix, each entry of which indicates the similarity between two points...

    Bin Tonget al. Subclass-Oriented Dimension Reduction with Constraint Transformation a...

    • ...Traditional answers include preservation of maximum variance as in principal component analysis (PCA); preservation of an independence structure as in independent component analysis; preservation of distances and pairwise constraints as in [1]; or maximization of class predictive power as in linear discriminant analysis, informative discriminant analysis [2], neighborhood components analysis [3], metric learning by collapsing classes [4], ...

    Jaakko Peltonen. Visualization by Linear Projections as Information Retrieval

    • ...Learning distance metrics is very important for various vision applications such as object classification, image retrieval, and video retrieval (Chen et al., 2005; Cevikalp et al., 2008; Hertz et al., 2003; Hadsell et al., 2006), and this task is much easier when the target values (labels) associated to the data samples are available...
    • ...Furthermore, learning an effective full rank distance metric by using side information cannot be carried out in such high-dimensional spaces since the number of parameters to be estimated is related to the square of the dimensionality and there is insufficient side information to obtain accurate estimates (Cevikalp et al., 2008)...
    • ...A better approach would be to use a semi-supervised dimensionality reduction method to find a low-dimensional embedding satisfying the pairwise equivalence constraints as in (Cevikalp et al., 2008)...
    • ...The proposed method bears similarity to the semi-supervised dimension reduction method introduced in (Cevikalp et al., 2008), but it does not assume that samples in a sufficiently small neighborhood tend to have same label...
    • ...The proposed Semi-Supervised Distance Metric Learning (SSDML) algorithm is compared to the full rank distance metric learning algorithm followed by dimensionality reduction and the Constrained Locality Preserving Projection (CLPP) method of (Cevikalp et al., 2008)...
    • ...As in (Cevikalp et al., 2008), we set the number of clusters to two, one cluster for the background and another for the object of interest...

    Hakan Cevikalpet al. Semi-supervised Distance Metric Learning for Visual Object Classificat...

Sort by: