In this paper, we propose an approach to learning appearance models of moving objects directly from compressed video. The appearance of a moving object changes dynamically in vide...
Our goal is to segment a video sequence into moving objects and the world scene. In recent work, spectral embedding of point trajectories based on 2D motion cues accumulated from ...
Human identity recognition is an important yet underaddressed
problem. Previous methods were strictly limited
to high quality photographs, where the principal techniques
heavily...
Periodicity is at the core of the recognition of many actions. This paper takes the following steps to detect and measure periodicity. 1) We establish a conceptual framework of cl...
Erik Pogalin, Arnold W. M. Smeulders, Andrew H. C....
We view the task of change detection as a problem of object recognition from learning. The object is defined in a 3D space where the time is the 3rd dimension. We propose two com...