Abstract. The aim of this paper is to present a new tool of multiple instance learning which is designed using a grammar based genetic programming (GGP) algorithm. We study its app...
This paper introduces a novel machine learning model called multiple instance ranking (MIRank) that enables ranking to be performed in a multiple instance learning setting. The mo...
Charles Bergeron, Jed Zaretzki, Curt M. Breneman, ...
We empirically study the relationship between supervised and multiple instance (MI) learning. Algorithms to learn various concepts have been adapted to the MI representation. Howe...
Multiple instance (MI) learning is a recent learning paradigm that is more flexible than standard supervised learning algorithms in the handling of label ambiguity. It has been u...
Abstract--Recently, classifier grids have shown to be a considerable alternative for object detection from static cameras. However, one drawback of such approaches is drifting if a...