Title :
Multi-view support vector machines for distributed activity recognition
Author :
Mosabbeb, Ehsan Adeli ; Raahemifar, Kaamran ; Fathy, Mahmood
Author_Institution :
Iran Univ. of Sci. & Technol., Tehran, Iran
Abstract :
In this paper, we propose a Multi-view Distributed SVM model. Most distributed classification models, distribute the instances among their processing nodes, while we assume that one instance is formed as a combination of information from different sources. This makes our model a great choice for multi-view activity recognition in camera sensor networks. We demonstrate the effectiveness of the algorithm, using the IXMAS dataset.
Keywords :
computer vision; image classification; object recognition; support vector machines; IXMAS dataset; camera sensor network; distributed activity recognition; distributed classification model; multiview activity recognition; multiview distributed SVM model; multiview support vector machine; Accuracy; Support vector machines; Vectors;
Conference_Titel :
Distributed Smart Cameras (ICDSC), 2013 Seventh International Conference on
Conference_Location :
Palm Springs, CA
DOI :
10.1109/ICDSC.2013.6778240
