Pattern Analysis with Graphs and Trees

author: Edwin Hancock, University of York

Description

Spectral representations of graphs, Pattern spaces from graph spectra, Spectral approaches to matching, Heat kernel methods Probabilistic and spectral methods for graph matching and clustering. Applications in computer vision.

Categories

Top: Computer Science: Machine Learning: Pattern Recognition

You might be experiencing some problems with Your Video player.

*Slides*
0:00	Constrained Delaunay Triangulation
0:55	Gabriel Graph
1:51	Relative Neighbourhood Graph
2:19	Shock graphs
4:17	Measuring the similarity of graphs
5:10	Starting point
5:22	Probabilistic modelling
7:15	Relational graph matching
8:05	Distribution of matching errors
14:10	Probability distribution for matching errors
15:57	Uses
17:18	Probability distribution for matching errors
17:31	Uses
17:33	Edit Distance
17:58	Literature
17:59	Spectral Correspondence Matching
19:24	Graph-spectral Methods for Correspondence
19:59	Umeyama’s Algorithm
22:18	Approach
23:07	Matrix Representation
24:30	Problem Formulation
26:23	Maximum Likelihood Framework
26:27	Factorial observation density
27:02	Bernoulli Distribution for Correspondences
29:09	Multiple Edge Constraints
29:12	Probability distribution for correspondences
30:39	Log-likelihood function
31:10	Expected log-likelihood
32:28	Probability distribution for correspondences
32:50	Expected log-likelihood
33:57	EM Algorithm
34:13	Maximisation Step
34:29	Matrix Representation
34:32	Expected log-likelihood
34:55	Matrix Representation
36:16	Maximisation using SVD
38:09	Expectation step
38:57	Maximisation using SVD
39:22	CMU House
39:38	Distortions
40:05	Luo – EM+SVD
40:23	Luo
40:28	Luo
40:34	Luo
40:38	Luo
40:40	Results Summary
41:12	Umeyama
41:23	Results Summary
41:33	Umeyama
41:57	Shapiro and Brady
42:00	Correspondences
42:40	Convergence
42:58	Edge Errors (Size constant)
43:21	Edge Errors versus Positional Jitter
43:24	Summary
44:13	Graph seriation

Lecture rating

People found this lecture:
Worth seeing
because it is:
Valuable and informative
Well presented
Easily understandable
Acceptably recorded
You need to login to cast your vote.

Report a problem or upload files

If you have found a problem with this lecture or would like to send us extra material, articles, exercises, etc., please use our ticket system to describe your request and upload the data.
Enter your e-mail into the 'Cc' field, and we will keep you updated with your request's status.