de
en
Schliessen
Detailsuche
Bibliotheken
Projekt
Impressum
Datenschutz
zum Inhalt
Detailsuche
Schnellsuche:
OK
Ergebnisliste
Titel
Titel
Inhalt
Inhalt
Seite
Seite
Im Dokument suchen
Kosov, Sergey: Multi-layer conditional random fields for revealing unobserved entitiesMehrschichtiges bedingtes Zufallsfeld zum Aufdecken nicht beobachteter Entitäten. 2018
Inhalt
Acknowledgements
Abstract
Zusammenfassung
Contents
Introduction
Motivation
Overview
Classification with Conditional Random Fields
Multi-Layered Conditional Random Fields
Related Work
Modeling of Conditional Random Fields
Modeling of Multi-Layer Conditional Random Fields
Organization and Contributions
Conditional Random Fields
Graphical Models
Directed Graphical Models | Bayesian Networks
Undirected Graphical Models | Markov Random Fields
Conditional Random Fields
Local-Global CRFs
Features
Global Features
DCNN Features
Confidence Features
Association Potentials
Generative Approaches
Naïve Bayes Model
Gaussian Models and Gaussian Mixture Models
Sequential Gaussian Mixture Model
Discriminative Approaches
K-Nearest Neighbors
Support Vector Machine
Random Forests
Artificial Neural Network
Interaction Potentials
Data Independent and Contrast Sensitive Approaches
Data Dependent Approaches
Histogram Matrix
Concatenated Edge Potentials
Inference and Decoding
Sum-Product Message Passing Algorithm
Max-Product Message Passing Algorithm
Parameter Estimation
Cross Validation Technique
Powell Search Method
Experiments
Impact of The Features on Classification
DCNN Features
Global Features
Confidence Features
Feature Importance Evaluation
Robust Association Potentials
Sequential GMM Model
Efficient KNN Model
Spatial Regularization
Local-Global CRF
Summary
Multi-Layer Conditional Random Fields
Occlusion Model
Depth Map Estimation
Mixed Graphical Models
The m-separation Criterion
Properties of m-connecting paths
Marginalization and Conditioning
Graph Transformation
Multi-Layer Graphical Models
Multi-Layer CRFs
Inter-Layer Pairwise Potential
Inference in Multi-Layer Graphs
Double Marginalization
Two-Layer Conditional Random Fields
Experiments
Two-Layer Conditional Random Fields
Multi-Layer Conditional Random Fields
Summary
Conclusion and Outlook
Conclusion
Future Work
Appendix Notations
Appendix Graphical Models
Statistical Independence
Conditional Independence
General Product Rule
Potential Approximation
Graph Construction Rules
Mixed Graphs
Appendix Algorithms
Function addPoint
Function distance
Function divergence
Appendix Datasets and Experimental Setup
Datasets
Reference Data
Features
Bibliography
Own Publications
Index