Stat 238

Vision as Bayesian Inference

Mon/Wed: 12:00-1:15 Winter 2014, Maths/Sciences 5203.


Course Description

This course models vision as Bayesian Inference. It concentrates on visual tasks such as segmenting images, detecting objects in images, and recognizing objects. Its goal is to describe the state of the art techniques. The handouts consist of copies of the lecture notes and related papers.

Reading Material

Grading Plan

3 homework assignments (25% each). Term project or review (25%).

Homework1

Tentative Schedule

Lecture Date Topics Handouts Supplements
Additional Readings
1 01-06 Introduction
Lecture1


2
01-08 Images, Linear Filters, and Statistical Edge Detection
Lecture2
Notes2
LinearFilteringKokkinos  KonishiPaper
3
01-13
Dictionaries, Super-Pixels, and More Edge Detection



3
01-15
Guest Lecture: Dr. Boyan Bonev. Hierarchical Super-pixels
Lecture4


4
01-20
Martin Luther King Holiday



5
01-22
Energy Functional Models
Lecture5

TotalVariationExample
6
01-27
Markov Random Fields: Gibbs Sampling, MFT
Lecture6
IntroGraphicalModels
MRFsMFTBPGibbsSamping
7
01-29
Belief Propagation
Lecture7
BeliefPropagation

8
02-03
Learning Distributions
Lecture8
HandNotes
Frame DellaPietra
9
02-05
Learning Distributions with Hidden Variables.
Lecture9
HiddenMarkovModels

10
02-10
Regression and Structural Support Vector Machines
Lecture9.5

YuilleHe
11
02-12
Lambertian Lighting Models
Lecture10
Basri1
Basri2
12
02-17
Preisdent's Day



13
02-19
The Geometry of Multiple Images Lecture11
Notes

14
02-24
Segmentation and Image Parsing Lecture12


15
02-26
AdaBoost/Regression: Face and Text detection Lecture13
ChenYuille

16
03-03
Deformable Template Models of Objects Lecture14


17
03-5
Deformable Part Models and PASCAL
Lecture15


18
03-10
Deep Neural Network Models



19
03-12
Hierarchical Compositional Models