Due to the increasing demands for network security, distributed intrusion detection has become a hot research topic in computer science. However, the design and maintenance of the...
We propose an integrated framework for automated hand model initialization and tracking using voxel data. Starting with an initial specific hand pose, the Laplacian Eigenspace (LE...
In this paper, we propose a probabilistic model for web image mining, which is based on concept-sensitive salient regions without human intervene. Our goal is to achieve a middle-...
Neural networks are a useful alternative to Gaussian mixture models for acoustic modeling; however, training multilayer networks involves a difficult, nonconvex optimization that...
A self-adaptive Hidden Markov Model (SA-HMM) based framework is proposed for behavior recognition in this paper. In this model, if an unknown sequence cannot be classified into an...