eduzhai > Applied Sciences > Engineering >

Independent Vector Analysis with Deep Neural Network Source Priors

  • king
  • (0) Download
  • 20210507
  • Save

... pages left unread,continue reading

Document pages: 5 pages

Abstract: This paper studies the density priors for independent vector analysis (IVA)with convolutive speech mixture separation as the exemplary application. Mostexisting source priors for IVA are too simplified to capture the finestructures of speeches. Here, we first time show that it is possible toefficiently estimate the derivative of speech density with universalapproximators like deep neural networks (DNN) by optimizing certain proxyseparation related performance indices. Experimental results suggest that theresultant neural network density priors consistently outperform previous onesin convergence speed for online implementation and signal-to-interference ratio(SIR) for batch implementation.

Please select stars to rate!


0 comments Sign in to leave a comment.

    Data loading, please wait...