The SRI speaker recognition system for the 2010 NIST speaker recognition evaluation (SRE) incorporates multiple subsystems with a variety of features and modeling techniques. We d...
Nicolas Scheffer, Luciana Ferrer, Martin Graciaren...
An auto-relation aided multiple-input/output inverse filtering algorithm (A-RAM) is proposed for the inverse filtering of room acoustics in speech dereverberation. In A-RAM, we ...
Abstract—The problem of minimizing queueing delay of opportunistic access of multiple continuous time Markov channels is considered. A new access policy based on myopic sensing a...
Most of the studies in the framework of Lambek calculus have considered the parsing process and ignored the generation process. This paper wants to rely on the close link between ...
Abstract—Putative brain processes responsible for understanding language are based on spreading activation in semantic networks, providing enhanced representations that involve c...