Time and frequency pruning for speaker identification

Abstract

This work is an attempt to refine decisions in speaker identification. A test utterance is divided into multiple time-frequency blocks on which a normalized likelihood score is calculated. Instead of averaging the block-likelihoods along the whole test utterance, some of them are rejected (pruning) and the final score is computed with a limited number of time-frequency blocks. The results obtained in the special case of time pruning lead the authors to experiment a joint time and frequency pruning approach. The optimal percentage of blocks pruned is learned on a tuning data set with the minimum identification error criterion. Validation of the time-frequency pruning process on 567 speakers leads to a significant error rate reduction (up to 41% reduction on TIMIT) for short training and test duration.