Post-Training Embedding Alignment for Decoupling Enrollment and Runtime Speaker Recognition Models

Open in new window