Boosting classification performance via data fusion
Costin Barbu, Jing Peng · 2015
An engine for fusing data from multiple sensors for classification is provided in this paper. Two novel methods for fusing multiple representations of data with boosting are presented and empirically evaluated against other fusion techniques as candidate algorithms for the fusion engine. We argue that information fusion from sensors operating in complementary regions of the spectrum and/or spatially separated can improve the classification performance.