Time synchronization of video sequences in a multi-camera system is necessary for successfully analyzing the acquired visual information. Even if synchronization is established, it...
Human behavior recognition is one of the most important and challenging objectives performed by intelligent vision systems. Several issues must be faced in this domain ranging fro...
Abstract. In speech recognition, phonemes have demonstrated their efficacy to model the words of a language. While they are well defined for languages, their extension to human act...
Kaustubh Kulkarni, Edmond Boyer, Radu Horaud, Amit...
Abstract. We consider a Newton-CG augmented Lagrangian method for solving semidefinite programming (SDP) problems from the perspective of approximate semismooth Newton methods. In ...
Abstract--This paper presents a new region-based unified tensor level set model for image segmentation. This model introduces a three-order tensor to comprehensively depict feature...