eTRIMS

eTRIMS - E-Training for Interpreting Images of Man-Made Scenes

eTRIMS

Tools:

Annotation Tool
Filip Korč, David Schneider

Dataset:

Benchmark data set of eTRIMS Image Database
Mar 31, 2009. Official release with 60 annotated images

Project Results
Phase III:

University of Bonn

	Learning Discriminative Markov Random Field Models for Image Interpretation Filip Korč
	The Role of Sequences for Incremental Learning Susanne Wenzel
	Hierarchical Conditional Random Field for Multi-class Image Classification Michael Ying Yang, Martin Drauschke

University of Hamburg

	Context-aware Classification Arne Kreutzmann, Kasim Terzić, Bernd Neuman
	Learning Probabilistic Structure Graphs Johannes Hartz
	Interpreting Structures in Man-Made Scenes: Combining Low-Level and High-Level Structure Sources Kasim Terzić, Lothar Hotz, Jan Šochman

Czech Technical University in Prague

Facade Image Parsing
Radim Šára, Jan Čech, Jan Šochman

Imperial College London

Fully Bottom up Interpretation of Man-Made Scenes
Mohammad Jahangiri, Daniel Heesch, Robby Tan and Maria Petrou

Project Results
Phase II:

University of Bonn

	Feature Selection for conditional Bayesian networks Martin Drauschke
	Semi-supervised incremental learning of hierarchical appearance models Susanne Wenzel, Wolfgang Förstner
	Probabilistic Reasoning for Context-Based Image Interpretation Filip Korč, Wolfgang Förstner

University of Hamburg

	Scene interpretation using learnt concepts
	Version Space Learning of spatial structures Johannes Hartz
	Middle layer in a scene interpretation system Kasim Terzić, Bernd Neumann
	Probabilistic models for compositional hierarchies Bernd Neumann

Czech Technical University in Prague

	Towards Facade Structure Recognition Jan Čech, Radim Šára
	Online Boosting for Scene Interpretation Jan Šochman

Imperial College London

Fully bottom up blob detection in building facades
Mohammad Jahangiri, Maria Petrou

Project Results
Phase I:

University of Bonn

	EOH-Based Classifier for the Interpretation of Urban Images Filip Korč
	Learning a Probabilistic Partonomy from Annotated Images Filip Korč
	Shape and Appearance Based Window Extraction Filip Korč
	Scale Space Hierarchy of Regions Martin Drauschke
	Detection of Repeated Structures in Facades Susanne Wenzel

University of Hamburg

	Probabilistic Models for Compositional Hierarchies Bernd Neumann
	Version Space Learning of Spatial Structures Johannes Hartz, Bernd Neumann

Czech Technical University in Prague

	Windowpane Detection based on MAP labeling Jan Čech, Radim Šára
	Learning and Detection of Higher-Level Primitives Using AdaBoost Jan ?ochman

Imperial College London

	Non-Gibbsian Markov Random Field for modelling spatial relationships in structured scenes Daniel Heesch, Maria Petrou
	Dictionary of neighbourhood configurations using fuzzy spatial and topological relations (work in progress)
	Extraction of regions of interest (work in progress)

HITEC

Scene Interpretation using Learnt Concepts
Lothar Hotz, Kasim Terzić

04.01.2010