Problem
Research questions and friction points this paper is trying to address.
muscle fatigue recognition
sEMG
Maximum Voluntary Contraction
signal variability
low SNR
FatigueFormer: Static-Temporal Feature Fusion for Robust sEMG-Based Muscle Fatigue Recognition
🤖 AI Summary
This study addresses the challenge of robust muscle fatigue recognition from surface electromyography (sEMG) signals, which suffer from high variability and low signal-to-noise ratio across different levels of maximum voluntary contraction (MVC). To this end, the authors propose FatigueFormer, a novel framework that decouples static and dynamic features via saliency guidance and models them separately using parallel Transformer encoders before fusing the representations to enhance cross-MVC generalization. The method innovatively integrates attention mechanisms to enable interpretable visualization of fatigue dynamics while supporting end-to-end training. Evaluated on a self-collected dataset comprising 30 subjects across four MVC levels (20%–80%), FatigueFormer achieves state-of-the-art accuracy and significantly improves the stability of mild fatigue detection and cross-condition generalization performance.
📝 Abstract
We present FatigueFormer, a semi-end-to-end framework that deliberately combines saliency-guided feature separation with deep temporal modeling to learn interpretable and generalizable muscle fatigue dynamics from surface electromyography (sEMG). Unlike prior approaches that struggle to maintain robustness across varying Maximum Voluntary Contraction (MVC) levels due to signal variability and low SNR, FatigueFormer employs parallel Transformer-based sequence encoders to separately capture static and temporal feature dynamics, fusing their complementary representations to improve performance stability across low- and high-MVC conditions. Evaluated on a self-collected dataset spanning 30 participants across four MVC levels (20-80%), it achieves state-of-the-art accuracy and strong generalization under mild-fatigue conditions. Beyond performance, FatigueFormer enables attention-based visualization of fatigue dynamics, revealing how feature groups and time windows contribute differently across varying MVC levels, offering interpretable insight into fatigue progression.
Innovation
Methods, ideas, or system contributions that make the work stand out.
FatigueFormer
static-temporal feature fusion
sEMG-based fatigue recognition
Transformer encoder
interpretable dynamics
🔎 Similar Papers
No similar papers found.
