signle test - Githubissues

ranchlai / speaker-verification

Speaker verification using ResnetSE (EER=0.0093) and ECAPA-TDNN

88 stars 21 forks source link

Open basaris opened 11 months ago

basaris commented 11 months ago

Can I test the model for only 2 .wav files in some way? How can I use the model to compare 2 speakers and not evaluating a whole database?