DepFlow: Disentangled Speech Generation to Mitigate Semantic Bias in Depression Detection

TL;DR

DepFlow is a novel speech synthesis framework that disentangles depression-related features to generate controllable, bias-mitigated speech data, improving depression detection robustness and enabling ethical clinical simulations.

Contribution

We introduce DepFlow, a three-stage disentangled speech generation model that controls depressive severity and mitigates semantic bias in depression detection datasets.

Findings

DepFlow achieves effective disentanglement with ROC-AUC of 0.693.

CDoA dataset improves depression detection macro-F1 by up to 12%.

DepFlow outperforms conventional augmentation in robustness and interpretability.

Abstract

Speech is a scalable and non-invasive biomarker for early mental health screening. However, widely used depression datasets like DAIC-WOZ exhibit strong coupling between linguistic sentiment and diagnostic labels, encouraging models to learn semantic shortcuts. As a result, model robustness may be compromised in real-world scenarios, such as Camouflaged Depression, where individuals maintain socially positive or neutral language despite underlying depressive states. To mitigate this semantic bias, we propose DepFlow, a three-stage depression-conditioned text-to-speech framework. First, a Depression Acoustic Encoder learns speaker- and content-invariant depression embeddings through adversarial training, achieving effective disentanglement while preserving depression discriminability (ROC-AUC: 0.693). Second, a flow-matching TTS model with FiLM modulation injects these embeddings into synthesis, enabling control over depressive severity while preserving content and speaker identity. Third, a prototype-based severity mapping mechanism provides smooth and interpretable manipulation across the depression continuum. Using DepFlow, we construct a Camouflage Depression-oriented Augmentation (CDoA) dataset that pairs depressed acoustic patterns with positive/neutral content from a sentiment-stratified text bank, creating acoustic-semantic mismatches underrepresented in natural data. Evaluated across three depression detection architectures, CDoA improves macro-F1 by 9%, 12%, and 5%, respectively, consistently outperforming conventional augmentation strategies in depression Detection. Beyond enhancing robustness, DepFlow provides a controllable synthesis platform for conversational systems and simulation-based evaluation, where real clinical data remains limited by ethical and coverage constraints.