en-de
en-es
en-fr
en-pt
en-sl
en
en-zh
0.25
0.5
0.75
1.25
1.5
1.75
2
Pattern Analysis with Graphs and Trees
Published on Feb 25, 20079094 Views
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. Applic
Related categories
Chapter list
Constrained Delaunay Triangulation00:00
Gabriel Graph00:55
Relative Neighbourhood Graph01:51
Shock graphs02:19
Measuring the similarity of graphs04:17
Starting point05:10
Probabilistic modelling 05:22
Relational graph matching 07:15
Distribution of matching errors08:05
Probability distribution for matching errors14:10
Uses15:57
Probability distribution for matching errors17:18
Uses17:31
Edit Distance 17:33
Literature17:58
Spectral Correspondence Matching17:59
Graph-spectral Methods for Correspondence19:24
Umeyama’s Algorithm19:59
Approach22:18
Matrix Representation23:07
Problem Formulation24:30
Maximum Likelihood Framework26:23
Factorial observation density 26:27
Bernoulli Distribution for Correspondences27:02
Multiple Edge Constraints29:09
Probability distribution for correspondences29:12
Log-likelihood function30:39
Expected log-likelihood31:10
Probability distribution for correspondences32:28
Expected log-likelihood32:50
EM Algorithm33:57
Maximisation Step 34:13
Matrix Representation34:29
Expected log-likelihood34:32
Matrix Representation34:55
Maximisation using SVD36:16
Expectation step38:09
Maximisation using SVD38:57
CMU House39:22
Distortions39:38
Luo – EM+SVD40:05
Luo40:23
Luo40:28
Luo40:34
Luo40:38
Results Summary40:40
Umeyama41:12
Results Summary41:23
Umeyama41:33
Shapiro and Brady41:57
Correspondences42:00
Convergence42:40
Edge Errors (Size constant)42:58
Edge Errors versus Positional Jitter43:21
Summary 43:24
Graph seriation44:13