I\textsuperscript{2}RiMA: Spectral Riemannian Representation with Temporal Attention for Mental Stress Detection based on EEG Signals

2026-07-01

AI Summary

This study addresses the challenges of cross-subject EEG-based mental stress detection, which are primarily hindered by substantial inter-individual variability and strong spectral specificity that conventional methods struggle to model effectively. To overcome these limitations, the authors propose a novel framework integrating spectrum-specific Riemannian geometry with adaptive temporal attention. Specifically, spatial covariance matrices are constructed at individual frequency bins and mapped onto the tangent space of symmetric positive definite (SPD) manifolds. A data-driven spectral clustering strategy then identifies the most informative frequency bands. Finally, an intra- and inter-segment attention mechanism jointly captures local spectral dynamics and global temporal context. Evaluated on three public datasets, the method consistently outperforms five state-of-the-art models, achieving a peak balanced accuracy of 82.78% with only 1.60M parameters and 31.95M FLOPs, thereby significantly enhancing both cross-subject generalization and computational efficiency.

Abstract

Cross-subject EEG stress detection remains challenging because discriminative stress-related patterns are both subject-dependent and frequency-specific. Conventional Riemannian methods model spatial covariance mainly in the time domain, overlooking neural oscillations that are critical for high-level cognitive state decoding, while standard temporal tokenization often fragments inter-slice temporal coherence. To address these limitations, we propose \method{}, an Intra-Inter Riemannian Manifold Attention Network for EEG-based stress detection. \method{} constructs spatial covariance matrices independently at each frequency point and maps them to the SPD tangent space, preserving channel-wise geometry together with frequency-specific discriminative cues. It further introduces frequency cluster aggregation to select informative spectral components and reduce redundancy by forming compact, data-driven frequency clusters aligned with EEG rhythms. Finally, an intra-inter slice attention module adaptively integrates local slice-level spectral dynamics and global temporal context across EEG sequences. Experiments on three datasets show that \method{} consistently outperforms five state-of-the-art baselines, achieving up to 82.78\% balanced accuracy while remaining efficient with only 1.60M parameters and 31.95M FLOPs.

Problem

• cross-subject
• EEG
• mental stress detection
• frequency-specific
• subject-dependent

Innovation

• Riemannian manifold
• spectral representation
• temporal attention
• frequency clustering
• EEG-based stress detection