# Curriculum learning for self-supervised speaker verification

arXiv: https://arxiv.org/abs/2203.14525 · HTML: https://mlsanity.com/paper/2203.14525/

## Key facts

- Cited by 7 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 28 Mar 2022.

## Built on by (most cited first, 7 of 7)

- [Self-Distillation Prototypes Network: Learning Robust Speaker Representations without Supervision](https://mlsanity.com/paper/2406.11169/) (2024)
- [Self-Supervised Frameworks for Speaker Verification via Bootstrapped Positive Sampling](https://mlsanity.com/paper/2501.17772/) (2025)
- [Pushing the Frontiers of Self-Distillation Prototypes Network with Dimension Regularization and Score Normalization](https://mlsanity.com/paper/2505.13826/) (2025)
- [SSPS: Self-Supervised Positive Sampling for Robust Self-Supervised Speaker Verification](https://mlsanity.com/paper/2505.14561/) (2025)
- [What and When to Learn: CURriculum Ranking Loss for Large-Scale Speaker Verification](https://mlsanity.com/paper/2603.24432/) (2026)
- [Self-Distillation Prototypes Network: Learning Robust Speaker Representations without Supervision](https://mlsanity.com/paper/2308.02774/) (2023)
- [Self-Supervised Learning for Speaker Recognition: A study and review](https://mlsanity.com/paper/2602.10829/) (2026)

## Abstract (arXiv)

The goal of this paper is to train effective self-supervised speaker representations without identity labels. We propose two curriculum learning strategies within a self-supervised learning framework. The first strategy aims to gradually increase the number of speakers in the training phase by enlarging the used portion of the train dataset. The second strategy applies various data augmentations to more utterances within a mini-batch as the training proceeds. A range of experiments conducted using the DINO self-supervised framework on the VoxCeleb1 evaluation protocol demonstrates the effectiveness of our proposed curriculum learning strategies. We report a competitive equal error rate of 4.47% with a single-phase training, and we also demonstrate that the performance further improves to 1.84% by fine-tuning on a small labelled dataset.