Leveraging In-the-Wild Data for Effective Self-Supervised Pretraining in Speaker Recognition

TL;DR

This paper demonstrates that self-supervised DINO training on large-scale in-the-wild speech data improves speaker recognition performance and introduces a confidence-based data filtering method for better pretraining data quality.

Contribution

It shows the effectiveness of DINO self-supervised learning on large in-the-wild datasets and proposes a confidence-based filtering algorithm to enhance pretraining data quality.

Findings

DINO training improves speaker recognition accuracy.

Confidence-based data filtering enhances pretraining efficiency.

Pretrained models and tools will be publicly available.

Abstract

Current speaker recognition systems primarily rely on supervised approaches, constrained by the scale of labeled datasets. To boost the system performance, researchers leverage large pretrained models such as WavLM to transfer learned high-level features to the downstream speaker recognition task. However, this approach introduces extra parameters as the pretrained model remains in the inference stage. Another group of researchers directly apply self-supervised methods such as DINO to speaker embedding learning, yet they have not explored its potential on large-scale in-the-wild datasets. In this paper, we present the effectiveness of DINO training on the large-scale WenetSpeech dataset and its transferability in enhancing the supervised system performance on the CNCeleb dataset. Additionally, we introduce a confidence-based data filtering algorithm to remove unreliable data from the pretraining dataset, leading to better performance with less training data. The associated pretrained models, confidence files, pretraining and finetuning scripts will be made available in the Wespeaker toolkit.