search

TaoRuijie / ECAPA-TDNN

Unofficial reimplementation of ECAPA-TDNN for speaker recognition (EER=0.86 for Vox1_O when train only in Vox2)
MIT License
581 stars 111 forks source link

When you trained the model on voxceleb, did you balanced the number of files per speaker ? #77

Open amitli1 opened 5 months ago

amitli1 commented 5 months ago
TaoRuijie commented 5 months ago

No, this imbalance is in a proper range.