Apple 发布半监督联邦 ASR 训练方案：在线伪标签配合服务器端更新稳定化
A Practical Recipe for Semi-Supervised Federated ASR: Online Pseudo-Labels with Server Update Stabilization
Apple 发了一篇联邦学习论文，讲怎么用伪标签在无标注数据上训 ASR 模型，还给了稳定训练的实操配方，做语音方向的可以看看。
Apple 机器学习团队发布一套半监督联邦学习（SSFL）方案，用于自动语音识别（ASR）训练。方法核心是客户端无标注数据由 teacher 模型生成伪标签，服务器端仅用小规模标注种子数据做 anchor 更新来稳定训练。论文指出 ASR 中伪标签错误会在输出序列和训练轮次间复合累积导致发散，因此将设计拆为 teacher 和 anchor 两个耦合轴。该方案旨在缩小与全监督联邦学习之间的差距。
A Practical Recipe for Semi-Supervised Federated ASR: Online Pseudo-Labels with Server Update Stabilization
Semi-supervised federated learning (SSFL) trains models on clients’ unlabeled data using a teacher to generate pseudo-labels, with a small labeled seed dataset on the server. Automatic Speech Recognition (ASR) is particularly fragile here: pseudo-label errors compound across the output sequence and across training rounds into divergence, leaving a large gap to fully-supervised FL. We show that closing this gap turns on two coupled design axes—the teacher (which model generates the pseudo-labels) and the anchor (the server-side updates on labeled data that stabilize training). On the teacher…
