In this paper, we propose a variational framework of finite Dirichlet mixture models and apply it to the challenging problem of object detection in static images. In our approach, the detection technique is based on the notion of visual keywords by learning models for object classes. Under the proposed variational framework, the parameters and the complexity of the Dirichlet mixture model can be estimated simultaneously, in a closed-form. The performance of the proposed method is tested on challenging real-world data sets. © 2011 Springer-Verlag.
CITATION STYLE
Fan, W., Bouguila, N., & Ziou, D. (2011). A variational statistical framework for object detection. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 7063 LNCS, pp. 276–283). https://doi.org/10.1007/978-3-642-24958-7_32
Mendeley helps you to discover research relevant for your work.