Attentive Statistics Pooling for Deep Speaker Embedding

TL;DR

This paper introduces attentive statistics pooling for deep speaker embedding, which uses an attention mechanism to weight frames differently and incorporate both mean and standard deviation, improving speaker verification accuracy.

Contribution

It presents a novel attentive pooling method that combines weighted means and standard deviations for more effective speaker feature representation.

Findings

Reduces EER by 7.5% on NIST SRE 2012

Reduces EER by 8.1% on VoxCeleb

Improves long-term speaker variation modeling

Abstract

This paper proposes attentive statistics pooling for deep speaker embedding in text-independent speaker verification. In conventional speaker embedding, frame-level features are averaged over all the frames of a single utterance to form an utterance-level feature. Our method utilizes an attention mechanism to give different weights to different frames and generates not only weighted means but also weighted standard deviations. In this way, it can capture long-term variations in speaker characteristics more effectively. An evaluation on the NIST SRE 2012 and the VoxCeleb data sets shows that it reduces equal error rates (EERs) from the conventional method by 7.5% and 8.1%, respectively.