M. Pawan Kumar
 
 

HOME

RESEARCH

PUBLICATIONS

GROUP

TALKS

TEACHING

CV

 

 

 

 

 

 

PART I (TAUGHT BY NIKOS KOMODAKIS)



SLIDES FOR PART II

Lecture 1: Reparameterization, belief propagation.
PPT   PDF

Lecture 2, Part 1: Convex relaxations.
PPT   PDF

Lecture 2, Part 2: Linear programming relaxation.
PPT   PDF

Lecture 3, Part 1: Tree-reweighted message passing.
PPT   PDF

Lecture 3, Part 2: Dual decomposition.
PPT   PDF

Lecture 4, Part 1: Rounding-based moves.
PPT   PDF

Lecture 4, Part 2: Sum-product belief propagation.
PPT   PDF

Lecture 4, Part 3: Minimizing free energy.
PPT   PDF

REFERENCES

The following paper provides a comparison of different convex relaxations.
An Analysis of Convex Relaxations for MAP Estimation of Discrete MRFs
by M. Pawan Kumar, Vladimir Kolmogorov and Phil Torr.

The following paper describes the sequential TRW algorithm. It also provides a brief description of belief propagation as reparameterization (subsection 2.2).
Convergent Tree-reweighted Message Passing for Energy Minimization
by Vladimir Kolmogorov.

The following paper describes the dual decomposition algorithm.
MRF Optimization via Dual Decomposition: Message-Passing Revisited
by Nikos Komodakis, Nikos Paragios and Georgios Tziritas.

The following paper presents an analysis of rounding-based move-making algorithms.
Rounding-based Moves for Metric Labeling
by M. Pawan Kumar.

PROGRAMMING ASSIGNMENT

Description

Data and Standard Code

FINAL EXAM FOR PART II

Model Exam Paper 1 (Difficult).

Model Exam Paper 2 (Easy).