Deep Normalization for Speaker Vectors

Abstract

Deep speaker embedding has demonstrated state-of-the-art performance in speaker recognition tasks. However, one potential issue with this approach is that the speaker vectors derived from deep embedding models tend to be non-Gaussian for each individual speaker, and non-homogeneous for distributions of different speakers. These irregular distributions can seriously impact speaker recognition performance, especially with the popular PLDA scoring method, which assumes homogeneous Gaussian distribution. In this paper, we argue that deep speaker vectors require deep normalization, and propose a deep normalization approach based on a novel discriminative normalization flow (DNF) model. We demonstrate the effectiveness of the proposed approach with experiments using the widely used SITW and CNCeleb corpora. In these experiments, the DNF-based normalization delivered substantial performance gains and also showed strong generalization capability in out-of-domain tests.

References in corpus (12)

• Adam: A Method for Stochastic Optimization
• VoxCeleb2: Deep Speaker Recognition
• VoxCeleb: a large-scale speaker identification dataset
• NICE: Non-linear Independent Components Estimation
• Density estimation using Real NVP
• Attentive Statistics Pooling for Deep Speaker Embedding
• Deep Speaker: an End-to-End Neural Speaker Embedding System
• Normalizing Flows for Probabilistic Modeling and Inference
• Guided Image Generation with Conditional Invertible Neural Networks
• BUT System Description to VoxCeleb Speaker Recognition Challenge 2019
• CN-CELEB: a challenging Chinese speaker recognition dataset
• VAE-based regularization for deep speaker embedding

Cited by in corpus (6)

• Remarks on Optimal Scores for Speaker Recognition
• Deep Speaker Vector Normalization with Maximum Gaussianality Training
• ASR-Free Pronunciation Assessment
• Neural Discriminant Analysis for Deep Speaker Embedding
• Deep generative LDA
• Deep generative factorization for speech signal