Self-supervised speech representation models have succeeded in various tasks, but improving them for content-related problems using unlabeled data is challenging. We propose speaker-invariant clustering (Spin), a novel self-supervised learning method that clusters speech representations and performs swapped prediction between the original and speaker-perturbed utterances. Spin disentangles speaker information and preserves content representations with just 45 minutes of fine-tuning on a single GPU. Spin improves pre-trained networks and outperforms prior methods in speech recognition and acoustic unit discovery.

翻译：自监督语音表示模型在多种任务中取得了成功，但利用无标签数据提升其处理内容相关问题的能力仍具挑战性。我们提出说话人不变量聚类（Spin），一种新颖的自监督学习方法，该方法对语音表示进行聚类，并在原始话语与经说话人扰动的话语之间执行交换预测。Spin仅需在单张GPU上微调45分钟，即可解耦说话人信息并保留内容表示。该模型可改进预训练网络，并在语音识别与声学单元发现任务中优于现有方法。