Name
Abstract | ||
|---|---|---|
This contribution treats the problem of learning andrecognizing 3D objects using 2D views. We present anew Bayesian approach to 3D computer vision basedon the Expectation--Maximization--Algorithm, wherelearning and classification of objects correspond to parameterestimation algorithms. We give a formal descriptionof different learning and recognition stagesand conclude the associated statistical optimizationproblems for each Bayesian decision. The trainingstage is supposed to be... |
Year | DOI | Venue |
|---|---|---|
1994 | 10.1109/ICPR.1994.577035 | ICPR |
Keywords | Field | DocType |
bayesian methods,image recognition,solid modeling,expectation maximization algorithm,optimization problem,random variables,bayesian approach,layout,normal distribution,density functional theory,distributed computing,image segmentation,parameter estimation,computer vision,object recognition | Mixture distribution,Observable,Computer science,Formal description,Artificial intelligence,Optimization problem,Special case,Computer vision,3D single-object recognition,Pattern recognition,Machine learning,Bayesian probability,Cognitive neuroscience of visual object recognition | Conference |
Citations | PageRank | References |
2 | 0.51 | 0 |
Authors | ||
2 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Joachim Hornegger | 1 | 168 | 17.44 |
| Heinrich Niemann | 2 | 1650 | 288.56 |