Kernel Methods in Computer Vision

By Christoph H. Lampert. 2009

137 citations recorded by the source corpus.

Over the last years, kernel methods have established themselves as powerful tools for computer vision researchers as well as for practitioners. In this tutorial, we give an introduction to kernel methods in computer vision from a geometric perspective, introducing not only the ubiquitous support vector machines, but also less known techniques for regression, dimensionality reduction, outlier detection, and clustering. Additionally, we give an outlook on very recent, non-classical techniques for the prediction of structure data, for the estimation of statistical dependency, and for learning the kernel function itself. All methods are illustrated with examples of successful application from the recent computer vision research literature.

Subfields: Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

Citations in the corpus

Transductive Multi-View Zero-Shot Learning (2015, methodology): “d the subsequent computation more robust. Speciﬁcally, we extend zeroscore normalisation to the similarity sets: (a) We assume 4. Most previous work [36], [56] sets $ by cross-validation. Inspired by [26], a simpler strategy for setting $is adopted: t median k;l=1; ;n i k ; j l > 2 in order to ha”
Attribute-Based Classification for Zero-Shot Visual Object Categorization (methodology): “…object classes, in particular faces, pedestrians, and vehicles, reliable and efficient detectors are available, based on the combination of powerful low-level features, such as SIFT [1] or HoG [2], with modern machine learning techniques, such as support vector machines [3], [4] or boosting [5].”
Random Laplace Feature Maps for Semigroup Kernels on Histograms (2014, background): “Despite its conceptual simplicity, this idea is highly generalizable and versatile: it leads to nonlinear algorithms for supervised image classification and object detection, unsupervised visual feature extraction, image denoising, action recognition in videos, integration of multiple descriptors [1”
Decision Forests: A Unified Framework for Classification, Regression, Density Estimation, Manifold Learning and Semi-Supervised Learning (2012, background): “In machine learning, the success of support vector classification has encouraged the development of support vector regression (SVR [57, 95]).”
Classifier adaptation at prediction time (2015, methodology): “SVMs are based on statistical learning theory and the maximum margin principle [14].”
Transductive Multi-view Embedding for Zero-Shot Recognition and Annotation (2014, unknown): “Inspired by [19], a rule of thumb for setting δ is δ ≈ median k,l=1,··· ,n < ψ k, ψ j l > 2 to balance roughly the same”
Decision Forests for Classification, Regression, Density Estimation, Manifold Learning and Semi-Supervised Learning (2011, background): “In machine learning, the success of support vector classification has encouraged the development of support vector regression (SVR [51, 86]).”
Stochastic Simulation of Patterns Using Distance-Based Pattern Modeling (2010, methodology): “In order to find the optimal bandwidth for our kernel function, we use a heuristic approach on the input data (Lampert, 2009).”