English

Discriminative Speaker Representation via Contrastive Learning with Class-Aware Attention in Angular Space

Audio and Speech Processing 2023-03-14 v3 Sound

Abstract

The challenges in applying contrastive learning to speaker verification (SV) are that the softmax-based contrastive loss lacks discriminative power and that the hard negative pairs can easily influence learning. To overcome the first challenge, we propose a contrastive learning SV framework incorporating an additive angular margin into the supervised contrastive loss in which the margin improves the speaker representation's discrimination ability. For the second challenge, we introduce a class-aware attention mechanism through which hard negative samples contribute less significantly to the supervised contrastive loss. We also employed gradient-based multi-objective optimization to balance the classification and contrastive loss. Experimental results on CN-Celeb and Voxceleb1 show that this new learning objective can cause the encoder to find an embedding space that exhibits great speaker discrimination across languages.

Keywords

Cite

@article{arxiv.2210.16622,
  title  = {Discriminative Speaker Representation via Contrastive Learning with Class-Aware Attention in Angular Space},
  author = {Zhe Li and Man-Wai Mak and Helen Mei-Ling Meng},
  journal= {arXiv preprint arXiv:2210.16622},
  year   = {2023}
}

Comments

Accepted by ICASSP 2023, 5 pages, 2 figures

R2 v1 2026-06-28T04:46:17.927Z