# Partial AUC optimization based deep speaker embeddings with class-center learning for text-independent speaker verification

arXiv: https://arxiv.org/abs/1911.08077 · HTML: https://mlsanity.com/paper/1911.08077/

## Key facts

- Cited by 5 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 19 Nov 2019.

## Built on by (most cited first, 5 of 5)

- [Learning with Multiclass AUC: Theory and Algorithms](https://mlsanity.com/paper/2107.13171/) (2021)
- [Optimizing Two-way Partial AUC with an End-to-end Framework](https://mlsanity.com/paper/2206.11655/) (2022)
- [Weakly Supervised AUC Optimization: A Unified Partial AUC Approach](https://mlsanity.com/paper/2305.14258/) (2023)
- [Generalizing AUC Optimization to Multiclass Classification for Audio Segmentation With Limited Training Data](https://mlsanity.com/paper/2110.14425/) (2021)
- [LMD: A Learnable Mask Network to Detect Adversarial Examples for Speaker Verification](https://mlsanity.com/paper/2211.00825/) (2022)

## Abstract (arXiv)

Deep embedding based text-independent speaker verification has demonstrated superior performance to traditional methods in many challenging scenarios. Its loss functions can be generally categorized into two classes, i.e., verification and identification. The verification loss functions match the pipeline of speaker verification, but their implementations are difficult. Thus, most state-of-the-art deep embedding methods use the identification loss functions with softmax output units or their variants. In this paper, we propose a verification loss function, named the maximization of partial area under the Receiver-operating-characteristic (ROC) curve (pAUC), for deep embedding based text-independent speaker verification. We also propose a class-center based training trial construction method to improve the training efficiency, which is critical for the proposed loss function to be comparable to the identification loss in performance. Experiments on the Speaker in the Wild (SITW) and NIST SRE 2016 datasets show that the proposed pAUC loss function is highly competitive with the state-of-the-art identification loss functions.