Memory Time Span in LSTMs for Multi-Speaker Source Separation.

Jeroen Zegers, Hugo Van hamme

Browse the full Interspeech paper archive.