We propose a new approach for singer melody extraction, based on blind source separation techniques. The short time Fourier transform (STFT) of the singer signal is modelled by a ...
In this paper we address the application of single sensor source separation techniques to mixtures of speech and music. Three strategies for source modeling are presented, namely ...
In this paper, we propose an MMSE a priori SNR estimator for speech enhancement. This estimator has similar benefits to the well-known decision-directed approach, but does not req...
— This paper provides a novel state vector and covariance sub-matrix recovery algorithm for a recently developed submap based exactly sparse Extended Information Filter (EIF) SLA...
This paper addresses nonclairvoyant and nonpreemptive online job scheduling in Grids. In the applied basic model, the Grid system consists of a large number of identical processor...
Uwe Schwiegelshohn, Andrei Tchernykh, Ramin Yahyap...