While most speech enhancement algorithms improve speech quality, they may not improve speech intelligibility in noise. This paper focuses on the development of an algorithm that ca...
Abstract--We consider inference in a general data-driven object-based model of multichannel audio data, assumed generated as a possibly underdetermined convolutive mixture of sourc...