逐层跨语言语音抑郁检测：基于对比对齐的分析

摘要

不同语言群体在抑郁症的诊断和临床表现在在显著差异。基于语音的抑郁症检测在单语环境下表现良好，但跨语言泛化仍是一个未解决的挑战。关键原因在于以往研究采用不区分说话人的片段级随机划分，导致身份泄露，从而夸大了报告指标。我们提出CLeaD——一种监督对比对齐框架，能够在不依赖平行数据或目标语言微调的情况下，将英语和普通话的WavLM嵌入映射到共享临床空间。对52名普通话说话者的评估显示，在留一说话人交叉验证下，对比对齐方法较基线略有提升（F1: 0.640 vs. 0.622），并在中间层（第7-8层）改善了抑郁类别的召回率，但较小的测试集限制了泛化能力。两个发现保持稳健：模型扩大会降低跨语言性能而提升单语英语性能；说话人身份泄露导致此前报道的普通话F1分数被人为膨胀至0.954——我们复现并量化了这一虚假现象。

English

Significant disparities exist in the diagnosis and clinical presentation of depression across different linguistic populations. Speech-based depression detection performs well monolingually, but cross-lingual generalization remains an open challenge. A key reason is that prior work uses segment-level random splits without speaker grouping, leading to identity leakage that inflates reported metrics. We propose CLeaD, a supervised contrastive alignment framework that maps WavLM embeddings from English and Mandarin into a shared clinical space, without parallel data or target-language fine-tuning. Evaluating 52 Mandarin speakers, contrastive alignment modestly outperforms the baseline (F1: 0.640 vs. 0.622) under leave-one-speaker-out evaluation. It also improves depressed-class recall at intermediate layers (7-8), though the small test set limits generalizability. Two findings remain robust: model scaling degrades cross-lingual performance while improving monolingual English, and speaker identity leakage artificially inflated previously reported Mandarin F1 scores to 0.954, an artifact we reproduce and quantify.