PRISM: Pre-trained Indeterminate Speaker Representation Model for Speaker Diarization and Speaker Verification.

Siqi Zheng, Hongbin Suo, Qian Chen

Browse the full Interspeech paper archive.