This paper presents a method for recognizing human actions based on pose primitives. In learning mode, the parameters representing poses and activities are estimated from videos. ...
Shadow detection is critical for robust and reliable vision-based systems for traffic vision analysis. Shadow points are often misclassified as object points causing errors in loc...
Background: A reliable extraction technique for resolving multiple spots in light or electron microscopic images is essential in investigations of the spatial distribution and dyn...
Local part-based human detectors are capable of handling partial occlusions efficiently and modeling shape articulations flexibly, while global shape template-based human detector...
Zhe Lin, Larry S. Davis, David S. Doermann, Daniel...
Abstract. The first step in various computer vision applications is a detection of moving objects. The prevalent pixel-wise models regard image pixels as independent random process...