This paper presents a generative model based approach to man-made structure detection in 2D natural images. The proposed approach uses a causal multiscale random field suggested i...
We present a Bayesian framework for learning higherorder transition models in video surveillance networks. Such higher-order models describe object movement between cameras in the...
Particle filtering algorithms can be used for the monitoring of dynamic systems with continuous state variables and without any constraints on the form of the probability distribu...
This paper addresses the problem of capturing the dynamics for exemplar-based recognition systems. Traditional HMM provides a probabilistic tool to capture system dynamics and in ...
Ahmed M. Elgammal, Vinay D. Shet, Yaser Yacoob, La...
Visual vocabulary serves as a fundamental component in many computer vision tasks, such as object recognition, visual search, and scene modeling. While state-of-the-art approaches...