# Cross attentive pooling for speaker verification

**Authors:** Seong Min Kye, Yoohwan Kwon, Joon Son Chung  
**arXiv:** [2008.05983](https://arxiv.org/abs/2008.05983)  
**Date:** 2020-08-13  
**Summary page:** https://www.summarizepaper.com/fr/paper/2008.05983v2/

## Abstract

The goal of this paper is text-independent speaker verification where utterances come from 'in the wild' videos and may contain irrelevant signal. While speaker verification is naturally a pair-wise problem, existing methods to produce the speaker embeddings are instance-wise. In this paper, we propose Cross Attentive Pooling (CAP) that utilizes the context information across the reference-query pair to generate utterance-level embeddings that contain the most discriminative information for the pair-wise matching problem. Experiments are performed on the VoxCeleb dataset in which our method outperforms comparable pooling strategies.