Adversarial Speaker Verification

Abstract

The use of deep networks to extract embeddings for speaker recognition has proven successfully. However, such embeddings are susceptible to performance degradation due to the mismatches among the training, enrollment, and test conditions. In this work, we propose an adversarial speaker verification (ASV) scheme to learn the condition-invariant deep embedding via adversarial multi-task training. In ASV, a speaker classification network and a condition identification network are jointly optimized to minimize the speaker classification loss and simultaneously mini-maximize the condition loss. The target labels of the condition network can be categorical (environment types) and continuous (SNR values). We further propose multi-factorial ASV to simultaneously suppress multiple factors that constitute the condition variability. Evaluated on a Microsoft Cortana text-dependent speaker verification task, the ASV achieves 8.8% and 14.5% relative improvements in equal error rates (EER) for known and unknown conditions, respectively.

5 pages, 1 figure, ICASSP 2019

References in corpus (10)

• Unsupervised Domain Adaptation by Backpropagation
• Deep Speaker: an End-to-End Neural Speaker Embedding System
• Parallel-Data-Free Voice Conversion Using Cycle-Consistent Adversarial Networks
• Speaker-Invariant Training via Adversarial Learning
• Adversarial Teacher-Student Learning for Unsupervised Domain Adaptation
• Unsupervised Adaptation with Domain Separation Networks for Robust Speech Recognition
• Invariant Representations for Noisy Speech Recognition
• Cycle-Consistent Speech Enhancement
• Adversarial Feature-Mapping for Speech Enhancement
• Adversarial Speaker Adaptation

Cited by in corpus (17)

• Speaker Adaptation for Attention-Based End-to-End Speech Recognition
• Adversarial Speaker Adaptation
• An Empirical Study on Channel Effects for Synthetic Voice Spoofing Countermeasure Systems
• Channel adversarial training for speaker verification and diarization
• Disentangled speaker and nuisance attribute embedding for robust speaker verification
• Attentive Adversarial Learning for Domain-Invariant Training
• An Adaptive X-vector Model for Text-independent Speaker Verification
• The INTERSPEECH 2020 Far-Field Speaker Verification Challenge
• Intra-class variation reduction of speaker representation in disentanglement framework
• Content-Aware Speaker Embeddings for Speaker Diarisation
• Data augmentation versus noise compensation for x- vector speaker recognition systems in noisy environments
• Active Voice Authentication
• SEEF-ALDR: A Speaker Embedding Enhancement Framework via Adversarial Learning based Disentangled Representation
• DEAAN: Disentangled Embedding and Adversarial Adaptation Network for Robust Speaker Representation Learning
• L-Vector: Neural Label Embedding for Domain Adaptation
• Within-sample variability-invariant loss for robust speaker recognition under noisy environments
• Multi-Level Transfer Learning from Near-Field to Far-Field Speaker Verification