In this paper we develop an algorithm for action recognition and localization in videos. The algorithm uses a figurecentric visual word representation. Different from previous ap...
Human movements are important cues for recognizing human actions, which can be captured by explicit modeling and tracking of actor or through space-time low-level features. Howeve...
: Web services over the Internet are widely used nowadays. The problem of secure access to Web-based systems is of great importance naturally. Compared with the existing models, th...
Abstract. This paper presents a new version of the CL contract specification language. CL combines deontic logic with propositional dynamic logic but it applies the modalities exc...
3D human pose estimation in multi-view settings benefits from embeddings of human actions in low-dimensional manifolds, but the complexity of the embeddings increases with the num...