Is it possible to eliminate the top-down model for speaker recognition with self-supervised speech features?
Self-supervised features are typically used instead of filterbank features in speech verification models. However, these models were originally designed to ...
