# FatigueFormer: A Physiology-Informed Model for Upper-Limb Muscle Fatigue Assessment

DevFeed: [FatigueFormer: A Physiology-Informed Model for Upper-Limb Muscle Fatigue Assessment](<https://devfeed.tech/articles/fatigueformer-a-physiology-informed-model-for-upper-limb-muscle-fatigue-assessment-116679.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2603.26841>)

Author: Tong Zhang, Hong Guo, Shuangzhou Yan, Dongkai Weng, Jian Wang, Hongxin Zhang

Published: 2026-11-04T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>), [arXiv: Machine Learning](<https://devfeed.tech/sources/arxiv-machine-learning.md>)

Topics: [Machine learning](<https://devfeed.tech/topics/machine-learning.md>), [Feature Engineering](<https://devfeed.tech/topics/feature-engineering.md>)

Tags: [assessment](<https://devfeed.tech/tags/assessment.md>), [context](<https://devfeed.tech/tags/context.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [dataset](<https://devfeed.tech/tags/dataset.md>), [deep-learning](<https://devfeed.tech/tags/deep-learning.md>), [encoding](<https://devfeed.tech/tags/encoding.md>), [external](<https://devfeed.tech/tags/external.md>), [feature](<https://devfeed.tech/tags/feature.md>), [leave](<https://devfeed.tech/tags/leave.md>), [machine](<https://devfeed.tech/tags/machine.md>), [machine-learning](<https://devfeed.tech/tags/machine-learning.md>), [model](<https://devfeed.tech/tags/model.md>), [modeling](<https://devfeed.tech/tags/modeling.md>), [mvc](<https://devfeed.tech/tags/mvc.md>), [offline](<https://devfeed.tech/tags/offline.md>), [patterns](<https://devfeed.tech/tags/patterns.md>), [performance](<https://devfeed.tech/tags/performance.md>), [properties](<https://devfeed.tech/tags/properties.md>), [recognition](<https://devfeed.tech/tags/recognition.md>), [state](<https://devfeed.tech/tags/state.md>), [surface](<https://devfeed.tech/tags/surface.md>)

## AI overview

FatigueFormer uses physiology-motivated grouping of sEMG descriptors into increasing- and decreasing-trend streams, combining context across windows with feature interactions within each window to recognize fatigue states defined by perceived exertion (RPE). The study introduces a dataset of 30 participants at four maximum voluntary contraction (MVC) levels and evaluates models offline using leave-one-subject-out and within-subject protocols, with external validation on SWIFTIES. Reported results support FatigueFormer across the examined MVC levels and both datasets, while suggesting fatigue cues vary with contraction intensity. An exploratory analysis found higher scores with more labeled target-user data on complete target recordings, including calibration samples.

## Source excerpt

arXiv:2603.26841v2 Announce Type: replace Abstract: Surface electromyographic (sEMG) signals provide information about muscle activity and function during movement. Fatigue alters multiple properties of these signals, with patterns that vary across individuals and contraction intensities. However, physiological insights into fatigue remain underused in sEMG modeling, particularly for cross-subject recognition. We introduce FatigueFormer to recognize fatigue states defined by ratings of perceived exertion (RPE). The model incorporates physiologically motivated trend grouping into sEMG encoding by organizing descriptors into increasing- and decreasing-trend streams and combining inter-window context with within-window feature interactions. To assess fatigue-state discriminability across contraction intensities, we further present an sEMG fatigue dataset collected from 30 participants at four levels of maximum voluntary contraction (MVC) and benchmark representative machine learning and deep learning methods. We evaluate performance offline under leave-one-subject-out (LOSO) and within-subject protocols and use the SWIFTIES dataset for external validation. Results support FatigueFormer's effectiveness across the examined contraction intensities and both datasets, while indicating that fatigue-related cues vary across MVC levels. An exploratory adaptation analysis shows higher FatigueFormer scores with more labeled target-user data on complete target recordings, which include calibration samples.