Print Email Facebook Twitter Intent-Aware Diverse Social Image Retrieval Title Intent-Aware Diverse Social Image Retrieval Author Wang, Bo (TU Delft Electrical Engineering, Mathematics and Computer Science; TU Delft Intelligent Systems) Contributor Larson, Martha (mentor) Degree granting institution Delft University of Technology Programme Computer Science Date 2018-11-21 Abstract Behind each photographic act is a rationale that impacts the visual appearanceof the resulting photo. Better understanding of this rationale has greatpotential to support image retrieval systems in serving user needs. However,at present, surprisingly little is known about the connection between whata picture shows (the literally depicted conceptual content) and why that picturewas taken (the photographer intent). In the thesis, we investigate photographerintent in a large Flickr data set. First, an expert annotator carriesout a large number of iterative intent judgments to create a taxonomy of intentclasses. Next, analysis of the distribution of concepts and intent classesreveals patterns of independence both at a global and user level. Finally,we report the results of experiments showing that a deep neural networkclassifier is capable of learning to differentiate between these intent classes,and that these classes support the diversification of image search results. Subject Multimedia retrievaluser intentmultimedia indexingdiversification To reference this document use: http://resolver.tudelft.nl/uuid:02324a33-32f5-49d6-9935-fa42dbe349d3 Part of collection Student theses Document type master thesis Rights © 2018 Bo Wang Files PDF bowang_thesis_final.pdf 4.97 MB Close viewer /islandora/object/uuid:02324a33-32f5-49d6-9935-fa42dbe349d3/datastream/OBJ/view