Multi-task Metric Learning For Text-independent Speaker Verification

Abstract

In this work, we introduce metric learning (ML) to enhance the deep embedding learning for text-independent speaker verification (SV). Specifically, the deep speaker embedding network is trained with conventional cross entropy loss and auxiliary pair-based ML loss function. For the auxiliary ML task, training samples of a mini-batch are first arranged into pairs, then positive and negative pairs are selected and weighted through their own and relative similarities, and finally the auxiliary ML loss is calculated by the similarity of the selected pairs. To evaluate the proposed method, we conduct experiments on the Speaker in the Wild (SITW) dataset. The results demonstrate the effectiveness of the proposed method.

Stats

• citations0
• S2 citations—
• github stars0
• HF likes0
• heat score0.00
• arxiv keychen2020multi

Related papers

• Improved Meta-learning Training For Speaker Verification (2021)4.52
• Centroid-based Deep Metric Learning For Speaker Recognition (2019)13.79
• A Comparison Of Metric Learning Loss Functions For End-to-end Speaker Verification (2020)6.77
• Masked Proxy Loss For Text-independent Speaker Verification (2020)2.26
• Multi-task Learning With High-order Statistics For X-vector Based Text-independent Speaker Verification (2019)8.35
• Partial AUC Optimization Based Deep Speaker Embeddings With Class-center Learning For Text-independent Speaker Verification (2019)9.59
• Metric Learning With Progressive Self-distillation For Audio-visual Embedding Learning (2025)3.58
• Improving The Gap In Visual Speech Recognition Between Normal And Silent Speech Based On Metric Learning (2023)0.00