— In this paper we propose a method that exploits 3D motion-based features between frames of 3D facial geometry sequences for dynamic facial expression recognition. An expressive...
Georgia Sandbach, Stefanos Zafeiriou, Maja Pantic,...
In this paper, we present a systematic framework for re-cognizing realistic actions from videos “in the wild.” Such unconstrained videos are abundant in personal collections as...
Jingen Liu (University of Central Florida), Jiebo ...
Images are highly complex multidimensional signals, with rich and complicated information content. For this reason they are difficult to analyze through a unique automated approach...
In the light of image retrieval evolving from text annotation to content-based and from standalone applications to web-based search engines, we foresee the need for deploying cont...
Extensive experimental evidence is required to study the impact of text categorization approaches on real data and to assess the performance within operational scenarios. In this ...
Roberto Basili, Alessandro Moschitti, Maria Teresa...