Jove
Visualize
Contact Us

Related Experiment Videos

Support vector machines for histogram-based image classification.

O Chapelle1, P Haffner, V N Vapnik

  • 1Speech and Image Processing Services Research Laboratory, AT&T Labs-Research, Red Bank, NJ 07701, USA.

IEEE Transactions on Neural Networks
|February 7, 2008
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

A European epidemiological survey of Vibrio splendidus clade shows unexplored diversity and massive exchange of virulence factors.

World journal of microbiology & biotechnology·2015
Same author

An overview of statistical learning theory.

IEEE transactions on neural networks·2008
Same author

Support vector machines for spam categorization.

IEEE transactions on neural networks·2008
Same author

Model complexity control for regression using VC generalization bounds.

IEEE transactions on neural networks·2008
Same author

Bounds on error expectation for support vector machines.

Neural computation·2000
Same author

[Retrospective evaluation and experimental radiographic study on bone reaction concomitant with orthodontic tooth movement].

Praktische Kieferorthopadie·1990
Same journal

Universal perceptron and DNA-like learning algorithm for binary neural networks: LSBF and PBF implementations.

IEEE transactions on neural networks·2013
Same journal

Guest editorial: special section on white box nonlinear prediction models.

IEEE transactions on neural networks·2011
Same journal

Data-based fault-tolerant control of high-speed trains with traction/braking notch nonlinearities and actuator failures.

IEEE transactions on neural networks·2011
Same journal

Guest editorial: special section on data-based control, modeling, and optimization.

IEEE transactions on neural networks·2011
Same journal

Neural network-based multiple robot simultaneous localization and mapping.

IEEE transactions on neural networks·2011
Same journal

Data-driven model-free adaptive control for a class of MIMO nonlinear discrete-time systems.

IEEE transactions on neural networks·2011
See all related articles
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Support vector machines (SVMs) with heavy-tailed radial basis function (RBF) kernels show improved image classification performance. A simple input remapping also enhances linear SVMs, offering a competitive alternative for high-dimensional histogram data.

Area of Science:

  • Computer Science
  • Machine Learning
  • Image Processing

Background:

  • Traditional classification methods struggle with high-dimensional feature spaces in image classification.
  • Generalization remains a challenge for existing approaches.

Purpose of the Study:

  • To evaluate the effectiveness of Support Vector Machines (SVMs) for image classification using high-dimensional histograms.
  • To investigate novel kernel functions and input transformations to improve SVM performance.

Main Methods:

  • Utilized Support Vector Machines (SVMs) with heavy-tailed Radial Basis Function (RBF) kernels of the form K(x, y) = e(-rho)Sigma(i)/xia-yia/b (a ≤ 1, b ≤ 2).
  • Tested on image classification using data from the Corel stock photo collection.
  • Compared performance against traditional polynomial and Gaussian RBF kernels.

Related Experiment Videos

  • Investigated the impact of input remapping (x(i)-->x(i)(a)) on linear SVMs.

    • Main Results:

    • Heavy-tailed RBF kernels significantly outperformed standard polynomial and Gaussian RBF kernels.
    • A simple input remapping technique notably improved the performance of linear SVMs.
    • Linear SVMs with input remapping became a viable alternative to RBF kernels for this task.

    Conclusions:

    • Heavy-tailed RBF kernels offer superior generalization for image classification with high-dimensional histogram features.
    • Input remapping presents a computationally efficient method to boost linear SVM performance.
    • SVMs, particularly with tailored kernels or input transformations, are effective for complex image classification problems.