Saved in:
Bibliographic Details
Main Authors: Shelke, Anushka Sanjay, Sneh, Aditya, Adyasha, Arya, Lone, Haroon R.
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2511.09039
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918198480207872
author Shelke, Anushka Sanjay
Sneh, Aditya
Adyasha, Arya
Lone, Haroon R.
author_facet Shelke, Anushka Sanjay
Sneh, Aditya
Adyasha, Arya
Lone, Haroon R.
contents Fairness in AI-driven stress detection is critical for equitable mental healthcare, yet existing models frequently exhibit gender bias, particularly in data-scarce scenarios. To address this, we propose FairM2S, a fairness-aware meta-learning framework for stress detection leveraging audio-visual data. FairM2S integrates Equalized Odds constraints during both meta-training and adaptation phases, employing adversarial gradient masking and fairness-constrained meta-updates to effectively mitigate bias. Evaluated against five state-of-the-art baselines, FairM2S achieves 78.1% accuracy while reducing the Equal Opportunity to 0.06, demonstrating substantial fairness gains. We also release SAVSD, a smartphone-captured dataset with gender annotations, designed to support fairness research in low-resource, real-world contexts. Together, these contributions position FairM2S as a state-of-the-art approach for equitable and scalable few-shot stress detection in mental health AI. We release our dataset and FairM2S publicly with this paper.
format Preprint
id arxiv_https___arxiv_org_abs_2511_09039
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Fairness-Aware Few-Shot Learning for Audio-Visual Stress Detection
Shelke, Anushka Sanjay
Sneh, Aditya
Adyasha, Arya
Lone, Haroon R.
Machine Learning
Computers and Society
Human-Computer Interaction
Fairness in AI-driven stress detection is critical for equitable mental healthcare, yet existing models frequently exhibit gender bias, particularly in data-scarce scenarios. To address this, we propose FairM2S, a fairness-aware meta-learning framework for stress detection leveraging audio-visual data. FairM2S integrates Equalized Odds constraints during both meta-training and adaptation phases, employing adversarial gradient masking and fairness-constrained meta-updates to effectively mitigate bias. Evaluated against five state-of-the-art baselines, FairM2S achieves 78.1% accuracy while reducing the Equal Opportunity to 0.06, demonstrating substantial fairness gains. We also release SAVSD, a smartphone-captured dataset with gender annotations, designed to support fairness research in low-resource, real-world contexts. Together, these contributions position FairM2S as a state-of-the-art approach for equitable and scalable few-shot stress detection in mental health AI. We release our dataset and FairM2S publicly with this paper.
title Fairness-Aware Few-Shot Learning for Audio-Visual Stress Detection
topic Machine Learning
Computers and Society
Human-Computer Interaction
url https://arxiv.org/abs/2511.09039