Skip to content

About

MelID is a ConvLSTM-based model for speaker ID/verification tasks, trained with Triplet Loss for embeddings generation.

Stars

0 stars

Watchers

1 watching

Forks

Releases

Packages

Contributors

Languages