Mitigating Gender Bias in Depression Detection via Counterfactual Inference

TL;DR

This paper introduces a counterfactual inference framework to reduce gender bias in audio-based depression detection models, improving fairness and accuracy by removing gender-related spurious correlations.

Contribution

The paper presents a novel causal inference-based debiasing method that effectively mitigates gender bias in depression detection models using counterfactual inference.

Findings

Significantly reduces gender bias in depression detection models.

Improves overall detection performance over existing debiasing methods.

Demonstrates effectiveness on the DAIC-WOZ dataset with acoustic backbones.

Abstract

Audio-based depression detection models have demonstrated promising performance but often suffer from gender bias due to imbalanced training data. Epidemiological statistics show a higher prevalence of depression in females, leading models to learn spurious correlations between gender and depression. Consequently, models tend to over-diagnose female patients while underperforming on male patients, raising significant fairness concerns. To address this, we propose a novel Counterfactual Debiasing Framework grounded in causal inference. We construct a causal graph to model the decision-making process and identify gender bias as the direct causal effect of gender on the prediction. During inference, we employ counterfactual inference to estimate and subtract this direct effect, ensuring the model relies primarily on authentic acoustic pathological features. Extensive experiments on the DAIC-WOZ dataset using two advanced acoustic backbones demonstrate that our framework not only significantly reduces gender bias but also improves overall detection performance compared to existing debiasing strategies.