Experimenting with Additive Margins for Contrastive Self-Supervised Speaker Verification

Open in new window