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