0000002737 19W 6SWS VI Computer Vision I: Variational Methods (IN2246)   Hilfe Logo

LV - Detailansicht

Wichtigste Meldungen anzeigenMeldungsfenster schließen
Allgemeine Angaben
Computer Vision I: Variational Methods (IN2246) 
lecture with integrated exercises
Winter semester 2019/20
Informatics 9 - Chair of Computer Vision and Artificial Intelligence (Prof. Cremers)
(Contact information)
Angaben zur Abhaltung
Variational Methods are among the most classical techniques for optimization of cost functions in higher dimension. Many challenges in Computer Vision and in other domains of research can be formulated as variational methods. Examples include denoising, deblurring, image segmentation, tracking, optical flow estimation, depth estimation from stereo images or 3D reconstruction from multiple views.

In this class, I will introduce the basic concepts of variational methods, the Euler-Lagrange calculus and partial differential equations. I will discuss how respective computer vision and image analysis challenges can be cast as variational problems and how they can be efficiently solved. Towards the end of the class, I will discuss convex formulations and convex relaxations which allow to compute optimal or near-optimal solutions in the variational setting.
The requirements for the class are knowledge in basic mathematics, in particular multivariate analysis and linear algebra. Some prior knowledge on optimization is a plus but is not necessary.
The aim of the class is to provide students with the ability of formulating Computer Vision challenges as variational problems and solving these by means of partial differential equations. The students should understand the concept of convexity and the advantages of convexity for optimization. And lastly the student should learn some strategies of how to convexify an originally non-convex problem.

Lecture, tutoarial (theory and programming), homework.
Excercise sheets will be passed out every week, containing problems for deepening the understanding of the topics covered in the course. In a weekly excercise group, we will discuss the solutions to the problems. Active participation in the excercises is the requirement for participation in the final exam. This will be written or oral, depending on the number of attendees. The questions will cover all material presented in class.
Für die Anmeldung zur Teilnahme müssen Sie sich in TUMonline als Studierende/r identifizieren.
Hints to the literature will be given throughout the lecture.