In this paper, we rst present a theorem that HOSVD-based representation of high-order tensor data provides a robust framework that can be used for a uni ed representation of the H...
In this paper we first introduce four kinds of modification of Symmetric Scoring [1] which produce likelihood ratios that do not need to be explicitly normalized, i.e. T-norm, Z...
People take more and more photos at different time and different events, however, these photos are often put into one giant folder and they are seldom annotated or organized. As t...
We have studied two efficient sampling methods, Langevin and Hessian adapted Metropolis Hastings (MH), applied to a parameter estimation problem of the mathematical model (Lorent...
The recently developed i-vector framework for speaker recognition has set a new performance standard in the research field. An i-vector is a compact representation of a speaker u...