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 ...
When building a classifier from clean training data for a particular test environment, knowledge about the environmental noise and channel should be taken into account. We propos...
Kevin Jamieson, Maya R. Gupta, Eric Swanson, Hyrum...
Abstract. This paper introduces the audio part of the 2010 communitybased Signal Separation Evaluation Campaign (SiSEC2010). Seven speech and music datasets were contributed, which...
We present an overview of the biomedical part of the 2010 community-based Signal Separation Evaluation Campaign (SiSEC2010), coordinated by the authors. In addition to the audio ta...
Shoko Araki, Fabian J. Theis, Guido Nolte, Dominik...
Signal degradation impacts the final quality of images acquired using remote sensing radiometer. The effectiveness of a restoration algorithm strongly depends on two main factors:...
Sebastiano Battiato, Giovanni Puglisi, Rosetta Riz...