# Analysis of ABC Frontend Audio Systems for the NIST-SRE24

arXiv: https://arxiv.org/abs/2505.15320 · HTML: https://mlsanity.com/paper/2505.15320/

## Key facts

- Cited by 3 papers in the map (data through 19 Oct 2026).
- First posted on arXiv 21 May 2025.

## Built on by (most cited first, 3 of 3)

- [TidyVoice: A Curated Multilingual Dataset for Speaker Verification Derived from Common Voice](https://mlsanity.com/paper/2601.16358/) (2026)
- [Xi+: Uncertainty Supervision for Robust Speaker Embedding](https://mlsanity.com/paper/2509.05993/) (2025)
- [U3-xi: Pushing the Boundaries of Speaker Recognition by Incorporating Uncertainty](https://mlsanity.com/paper/2601.15719/) (2026)

## Abstract (arXiv)

We present a comprehensive analysis of the embedding extractors (frontends) developed by the ABC team for the audio track of NIST SRE 2024. We follow the two scenarios imposed by NIST: using only a provided set of telephone recordings for training (fixed) or adding publicly available data (open condition). Under these constraints, we develop the best possible speaker embedding extractors for the pre-dominant conversational telephone speech (CTS) domain. We explored architectures based on ResNet with different pooling mechanisms, recently introduced ReDimNet architecture, as well as a system based on the XLS-R model, which represents the family of large pre-trained self-supervised models. In open condition, we train on VoxBlink2 dataset, containing 110 thousand speakers across multiple languages. We observed a good performance and robustness of VoxBlink-trained models, and our experiments show practical recipes for developing state-of-the-art frontends for speaker recognition.