In the 90s, a new type of learning algorithm was designed, based on results from statistical learning theory: the Support Vector Machine (SVM). This gave rise to the development of a new class of theoretically elegant learning machines which use a central concept of SVMs --- \emph{kernels} --- for a number of different learning tasks. Kernel machines now provide a modular and simple to use framework that can be adapted to different tasks and domains by the choice of the kernel function and the base algorithm, and they have been shown to perform very well in a wide range of problems. The talk will describe the basic ideas as well as some recent applications from computer vision and computer graphics.