Images as bags of pixels

Jebara,

doi:10.1109/iccv.2003.1238352

Cited by 44 publications

(33 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The work of [3] investigated image classification with kernels on color histograms. The bagof-pixels kernels proposed in [15,6] compare color distribution of two images by using kernels between probability measures, and was extended to hierarchical multi-scale settings in [6,18]. A different approach, taking into account the non-exchangeability of pixels in an image, is the one of [25] where the edit distance between two graphs is directly plugged into a kernel to defined a similarity measure between graphs.…”

Section: Previous Workmentioning

confidence: 99%

Image Classification with Segmentation Graph Kernels

Harchaoui

Bach

2007

2007 IEEE Conference on Computer Vision and Pattern Recognition

199

165

View full text Add to dashboard Cite

show abstract

Section: Previous Workmentioning

confidence: 99%

Image Classification with Segmentation Graph Kernels

Harchaoui

Bach

2007

2007 IEEE Conference on Computer Vision and Pattern Recognition

199

165

View full text Add to dashboard Cite

show abstract

“…tackle this problem but suffer from the usual sensitivity to the initialization, exhibiting uncertain convergence. When the relative orientation of the shapes to compare is known, the estimation of the permutation relating the point sets can be casted into a convex optimization problem [5]. However, normalizing a point set in what respects to rotation is harder than it could seem at first sight.…”

Section: Introductionmentioning

confidence: 99%

Shape representation via elementary symmetric polynomials: A complete invariant inspired by the bispectrum

Negrinho¹,

Aguiar²

2013

2013 IEEE International Conference on Image Processing

View full text Add to dashboard Cite

We address the representation of two-dimensional shape in its most general form, i.e., arbitrary sets of points, that may arise in multiple situations, e.g., sparse sets of specific landmarks, or dense sets of image edge points. Our goal are recognition tasks, where the key is balancing two contradicting demands: shapes that differ by rigid transformations or point re-labeling should have the same representation (invariance) but geometrically distinct shapes should have different representations (completeness). In the paper, we introduce a new shape representation that marries properties of the elementary symmetric polynomials and the bispectrum. Like the power spectrum, the bispectrum is insensitive to signal shifts; however, unlike the power spectrum, the bispectrum is complete. The elementary symmetric polynomials are complete and invariant to variable relabeling. We show that the elementary symmetric polynomials of the shape points depend on the shape orientation in a way that enables interpreting them in the frequency domain and building from them a bispectrum. The result is a shape representation that is complete and invariant to rigid transformations and point-relabeling. The paper also reports experiments that illustrate the proved properties.

show abstract

“…For instance, in document categorization, documents are usually represented as "bag-of-words", which are unordered set of key words. Images can also be treated as "bag-of-tuples", where the element is the tuple of the position and intensity of a pixel in an image [8]. Instead of directly defining a kernel between two sets, we take the methodology of first modeling each set probabilistically, and then constructing a kernel between the two probabilistic models.…”

Section: Introductionmentioning

confidence: 99%

A Kernel Between Unordered Sets of Data: The Gaussian Mixture Approach

Lyu

2005

Machine Learning: ECML 2005

View full text Add to dashboard Cite

Abstract. In this paper, we present a new kernel for unordered sets of data of the same type. It works by first fitting a set with a Gaussian mixture, then evaluate an efficient kernel on the two fitted Gaussian mixtures. Furthermore, we show that this kernel can be extended to sets embedded in a feature space implicitly defined by another kernel, where Gaussian mixtures are fitted with the kernelized EM algorithm [6], and the kernel for Gaussian mixtures are modified to use the outputs from the kernelized EM. All computation depends on data only through their inner products as evaluations of the base kernel. The kernel is computable in closed form, and being able to work in a feature space improves its flexibility and applicability. Its performance is evaluated in experiments on both synthesized and real data.

show abstract

Images as bags of pixels

Cited by 44 publications

References 10 publications

Image Classification with Segmentation Graph Kernels

Image Classification with Segmentation Graph Kernels

Shape representation via elementary symmetric polynomials: A complete invariant inspired by the bispectrum

A Kernel Between Unordered Sets of Data: The Gaussian Mixture Approach

Contact Info

Product

Resources

About