MeanShift, Segmentation as Energy Minimization
Kapitel:
Start  Kapitel 

00:00:00

Lecture 7: Segmentation as Energy Minimization 
00:00:09

Announcements 
00:01:16

Course Outline 
00:02:09

Recap: Image Segmentation 
00:02:57

Recap: KMeans Clustering 
00:05:16

Recap: Expectation Maximization (EM) 
00:07:51

Recap: EM Algorithm 
00:11:18

MoG Color Models for Image Segmentation 
00:13:38

Finding Modes in a Histogram 
00:16:42

MeanShift Segmentation 
00:17:12

MeanShift Algorithm 
00:18:44

MeanShift 
00:22:18

Real Modality Analysis 
00:22:59

MeanShift Clustering 
00:24:55

MeanShift Clustering/Segmentation 
00:26:58

MeanShift Segmentation Results 
00:27:21

More Results (1) 
00:27:59

More Results (2) 
00:28:05

Problem: Computational Complexity 
00:28:47

Speedups (1): Basin of Attraction 
00:29:26

Speedups (2) 
00:33:00

Summary MeanShift 
00:39:47

Back to the Image Segmentation Problem... 
00:40:54

Topics of This Lecture 
00:41:17

Markov Random Fields 
00:43:24

MRF Nodes as Pixels 
00:45:09

Network Joint Probability 
00:47:04

Energy Formulation (1) 
00:50:15

Energy Formulation (2) 
00:53:54

Energy Minimization 
00:55:45

Topics of This Lecture 
00:55:51

Graph Cuts for Optimal Boundary Detection 
00:59:24

Simple Example of Energy 
01:01:46

Adding Regional Properties (1) 
01:04:00

Adding Regional Properties (2) 
01:04:33

Adding Regional Properties (3) 
01:06:02

How to Set the Potentials? Some Examples 
01:09:29

Example: MRF for Image Segmentation 
01:14:16

Topics of This Lecture 
01:14:25

How Does it Work? The stMincut Problem 
01:14:41

The stMincut Problem 
01:15:24

What is the stmincut? 
01:15:37

How to Compute the stMincut? 
01:16:47

History of Maxflow Algorithms 
01:19:12

Maxflow Algorithms 
01:21:34

Applications: Maxflow in Computer Vision 
01:22:22

When Can st Graph Cuts Be Applied? 
01:25:16

Topics of This Lecture 
01:25:21

Dealing with NonBinary Cases 
01:25:49

αExpansion Move 
01:26:22

αExpansion Algorithm 
01:26:43

References and Further Reading 