1. Clustering-Based Analysis of Semantic Concept Models for Video Shots
- Author
-
Markus Koskela and Alan F. Smeaton
- Subjects
Signal processing ,Information retrieval ,Computer science ,Digital video ,Feature extraction ,Ontology (information science) ,Lexicon ,TRECVID ,Weighting ,Annotation ,Semantic similarity ,Semantic computing ,Entropy (information theory) ,Cluster analysis ,Multimedia systems - Abstract
In this paper we present a clustering-based method for representing semantic concepts on multimodal low-level feature spaces and study the evaluation of the goodness of such models with entropy-based methods. As different semantic concepts in video are most accurately represented with different features and modalities, we utilize the relative model-wise confidence values of the feature extraction techniques in weighting them automatically. The method also provides a natural way of measuring the similarity of different concepts in a multimedia lexicon. The experiments of the paper are conducted using the development set of the TRECVID 2005 corpus together with a common annotation for 39 semantic concepts.
- Published
- 2006