In this paper we propose a new technique to enhance emotion recognition by combining in different ways what we call emotion predictions. The technique is called F2 as the combinat...
We propose a new algorithm for fusion transformation that allows both stacks and accumulating parameters. The new algorithm can fuse programs that cannot be handled by existing fu...
Abstract. Two difficulties in designing data-centric routes [2–5] in wireless sensor networks are the lack of reasonably practical data aggregation models and the high computatio...
Semantic analysis of multimodal video aims to index segments of interest at a conceptual level. In reaching this goal, it requires an analysis of several information streams. At s...
Cees Snoek, Marcel Worring, Arnold W. M. Smeulders
Abstract. Most previous methods for generic object recognition explicitly or implicitly assume that an image contains objects from a single category, although objects from multiple...
Takahiro Okabe, Yuhi Kondo, Kris M. Kitani, Yoichi...