ASAIOct 18, 2023

DASA: Difficulty-Aware Semantic Augmentation for Speaker Verification

arXiv:2310.12111v11 citationsh-index: 20
Originality Incremental advance
AI Analysis

This work addresses the need for efficient and diverse data augmentation in speaker verification systems, offering an incremental improvement over existing methods.

The paper tackles the problem of limited diversity and high computational cost in data augmentation for speaker verification by introducing a difficulty-aware semantic augmentation (DASA) approach that generates diversified samples in embedding space with minimal extra cost, resulting in a 14.6% relative reduction in EER on the CN-Celeb evaluation set.

Data augmentation is vital to the generalization ability and robustness of deep neural networks (DNNs) models. Existing augmentation methods for speaker verification manipulate the raw signal, which are time-consuming and the augmented samples lack diversity. In this paper, we present a novel difficulty-aware semantic augmentation (DASA) approach for speaker verification, which can generate diversified training samples in speaker embedding space with negligible extra computing cost. Firstly, we augment training samples by perturbing speaker embeddings along semantic directions, which are obtained from speaker-wise covariance matrices. Secondly, accurate covariance matrices are estimated from robust speaker embeddings during training, so we introduce difficultyaware additive margin softmax (DAAM-Softmax) to obtain optimal speaker embeddings. Finally, we assume the number of augmented samples goes to infinity and derive a closed-form upper bound of the expected loss with DASA, which achieves compatibility and efficiency. Extensive experiments demonstrate the proposed approach can achieve a remarkable performance improvement. The best result achieves a 14.6% relative reduction in EER metric on CN-Celeb evaluation set.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes