Bidirectional Multiscale Feature Aggregation for Speaker Verification

Abstract

In this paper, we propose a novel bidirectional multiscale feature aggregation (BMFA) network with attentional fusion modules for text-independent speaker verification. The feature maps from different stages of the backbone network are iteratively combined and refined in both a bottom-up and top-down manner. Furthermore, instead of simple concatenation or element-wise addition of feature maps from different stages, an attentional fusion module is designed to compute the fusion weights. Experiments are conducted on the NIST SRE16 and VoxCeleb1 datasets. The experimental results demonstrate the effectiveness of the bidirectional aggregation strategy and show that the proposed attentional fusion module can further improve the performance.

References in corpus (7)

• Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
• VoxCeleb2: Deep Speaker Recognition
• ECAPA-TDNN: Emphasized Channel Attention, Propagation and Aggregation in TDNN Based Speaker Verification
• SphereFace: Deep Hypersphere Embedding for Face Recognition
• Improving Multi-Scale Aggregation Using Feature Pyramid Module for Robust Speaker Verification of Variable-Duration Utterances
• Exploring the Encoding Layer and Loss Function in End-to-End Speaker and Language Recognition System
• Spatial Pyramid Encoding with Convex Length Normalization for Text-Independent Speaker Verification