Probabilistic Graphical Models in Computer Vision (IN2329) (2h + 2h, 5 ECTS)
Several problems in computer vision can be cast as a labeling problem. Typically, such problems arise from Markov Random Field (MRF) models, which provide an elegant framework of formulating various types of labeling problems in imaging.
By making use of certain assumptions some „nice“ MRF models can be solved in polynomial time, whereas others are NP hard. We will see both, efficient algorithms for solving the „nice“ problems and relaxation strategies for the „hard“ ones.
The following topics will be covered in this module:
Directed and undirected graphical models
- Bayesian network
- Markov random field
- Conditional random field
Parameter learning for MRF and CRF models
- Gradient based optimization
- Stochastic gradient descent
- Structured support vector machine
Exact MAP inference methods for MRFs
- Belief propagation on trees: Max-sum algorithm
- Binary graph cuts
Approximate inference methods
- Loopy belief propagation
- Mean field approximation
- Graph cuts: alpha expansion, alpha-beta swap
- Linear programming relaxations: fast primal-dual schema
Practical applications that we will cover are:
- Binary and multi-label image segmentation
- Human pose estimation
- Stereo matching
- Object detection
The course is intended for Master students.
The requirements for the class are knowledge in basic mathematics, in particular multivariate analysis and linear algebra, and in basic computer science, in particular programming skills (eg.,
Matlab) and algorithms and data structures (eg., dynamic programming).
Please do not forget to register to the exam in the TUM online.
The exam will be oral.
The time slots for the exams are announced on the internal page.
Location: Room 02.09.023
Time and Date: Monday 10:15 - 12:00 (changed)
Lecturer: Dr. Csaba Domokos
Start: 24 April 2017
The lectures are held in English.
|Date of Lecture||Printer Friendly Lecture Slides|
|April, 24th||01. Introduction|
|May, 8th||02. Graphical models|
|May, 15th||03. Conditional random field, Expectation-maximization algorithm|
|May, 22nd||04. Mixture of Gaussians, Graph cut|
|May, 29th||05. Move making algorithms|
|June, 12th||06. Alpha-expansion, Primal-dual scheme|
|June, 19th||07. FastPD, Branch-and-MinCut|
|June, 26th||08. Belief propagation|
|July, 3rd||09. Human pose estimation, Mean field approximation|
|July, 10th||10. Sampling, Parameter learning|
|July, 17th||11. Parameter learning|
Location: Room 02.05.014
Time and Date: Monday 8:15 - 10:00 (changed)
Start: 24 April 2017
|Date of Tutorial||Exercise Sheet|
|April, 24th||Exercise sheet 1|
|May, 8th||Exercise sheet 2|
|May, 15th||Exercise sheet 3|
|May, 22nd||Exercise sheet 4|
|May, 29th||Exercise sheet 5|
|June, 12th||Exercise sheet 6|
|June, 19th||Exercise sheet 7|
|June, 26th||Exercise sheet 8|
|July, 3rd||Exercise sheet 9|
|July, 10th||Exercise sheet 10|
|July, 17th||Exercise sheet 11|
|July, 31st||Exercise sheet 12|
Solution sheets and extra announcements can be accessed here.
The password will be presented in the first lecture.
- D. Koller, N. Friedman. Probabilistic Graphical Models: Principles and Techniques, MIT Press, 2009.
- S. Nowozin, C. H. Lampert. Structured Learning and Prediction in Computer Vision, Foundations and Trends in Computer Graphics and Vision, 2011. Download
- A. Blake, P. Kohli, C. Rother. Markov Random Fields for Vision and Image Processing, MIT Press, 2011.