de
en
Schliessen
Detailsuche
Bibliotheken
Projekt
Impressum
Datenschutz
zum Inhalt
Detailsuche
Schnellsuche:
OK
Ergebnisliste
Titel
Titel
Inhalt
Inhalt
Seite
Seite
Im Dokument suchen
Struwe, Marvin: Active occlusion-handling for appearance-based object recognition models. 2017
Inhalt
Introduction
Scope of this Manuscript
Publications in the Context of this Thesis
Related Work
Introduction
Object Representation
Holistic Discriminative Approaches
Parts-Based Approaches
Object Relation
Global Context
Motion Features
Analytic Feature Framework
Introduction
Analytic Feature Framework
Adaptation of the Analytic Feature Framework to Car Detection
Datasets
Detection Results
Conclusion
Use of Object-Object Relations with a Holistic Discriminative Classifier
Introduction
Occlusion-handling Using Object-Object Relations
Datasets
Training of the Car Classifier
Detection Results
Conclusion
Split of the Holistic Car Model
Introduction
Split of the Holistic Car Template
Additional Use of Depth Information
Detection Examples
Conclusion
Rendered Benchmark Dataset
Introduction
Rendered Benchmark Dataset
Occlusion-handling Strategies of a Parts-based Car Detector
Introduction
Parts-based Car Detector
Extraction of Texture Descriptors
Learning of Parts-based Object Representation
Parts-based Detection Framework
Analysis of the Detection Performance and Computation Time of the Code-book
Occlusion-handling of the Parts-based Car Detector
Contribution-aware Strategy for Occlusion-handling of a Parts-based Car Detector
Detection Examples
Conclusion
Rotation and Illumination Stability of Texture Descriptors
Introduction
Datasets
Proof of Rotation and Illumination Robustness of Texture Descriptors
Discriminative Features
Conclusion
Psychophysical Study on Object Detection of Occluded Objects
Occluded Object Recognition by Humans
Dataset
Experimental Setup
Physical
Participants
Test Procedure
Evaluation
Conclusion
Conclusion