Toward Improving Synthetic Audio Spoofing Detection Robustness via Meta-Learning and Disentangled Training With Adversarial Examples

TL;DR

This paper introduces a robust spoofing detection system for automatic speaker verification that leverages meta-learning, adversarial data augmentation, and attention mechanisms to improve detection of unseen spoofing attacks.

Contribution

It proposes a novel combination of weighted loss, meta-learning, adversarial examples, and attention modules to enhance spoofing detection robustness against unseen attacks.

Findings

Achieved a pooled EER of 0.87% on ASVspoof 2019 LA dataset.

Reduced min t-DCF to 0.0277, indicating improved detection performance.

Demonstrated effectiveness of adversarial augmentation and meta-learning in spoofing detection.

Abstract

Advances in automatic speaker verification (ASV) promote research into the formulation of spoofing detection systems for real-world applications. The performance of ASV systems can be degraded severely by multiple types of spoofing attacks, namely, synthetic speech (SS), voice conversion (VC), replay, twins and impersonation, especially in the case of unseen synthetic spoofing attacks. A reliable and robust spoofing detection system can act as a security gate to filter out spoofing attacks instead of having them reach the ASV system. A weighted additive angular margin loss is proposed to address the data imbalance issue, and different margins has been assigned to improve generalization to unseen spoofing attacks in this study. Meanwhile, we incorporate a meta-learning loss function to optimize differences between the embeddings of support versus query set in order to learn a spoofing-category-independent embedding space for utterances. Furthermore, we craft adversarial examples by adding imperceptible perturbations to spoofing speech as a data augmentation strategy, then we use an auxiliary batch normalization (BN) to guarantee that corresponding normalization statistics are performed exclusively on the adversarial examples. Additionally, A simple attention module is integrated into the residual block to refine the feature extraction process. Evaluation results on the Logical Access (LA) track of the ASVspoof 2019 corpus provides confirmation of our proposed approaches' effectiveness in terms of a pooled EER of 0.87%, and a min t-DCF of 0.0277. These advancements offer effective options to reduce the impact of spoofing attacks on voice recognition/authentication systems.