Presentations (Communicative Events)

Using Source Models in Speech Separation

Ellis, Daniel P. W.

Discusses work on using ASR models to recognize mixtures and recovering spatial information in reverb.


  • thumnail for radcliffe-models-2007-11.pdf radcliffe-models-2007-11.pdf application/pdf 2.97 MB Download File

More About This Work

Academic Units
Electrical Engineering
Published Here
July 13, 2012


Presented at the Next-Generation Statistical Models and Inference for Speech and Audio Signal Processing seminar, Radcliffe Institute for Advanced Study, Cambridge, Mass., November 9, 2007.