Towards Detecting Biceps Muscle Fatigue in Gym Activity Using Wearables.

摘要: Fatigue is a naturally occurring phenomenon during human activities, but it poses bigger risk for injuries physically demanding such as gym activities and athletics. Several studies show that bicep muscle fatigue can lead to various may require up 22 weeks of treatment. In this work, we adopt wearable approach detect biceps concentration curl exercise an example activity. Our dataset consists 3000 curls from twenty middle-aged volunteers at ages between 27 30 Body Mass Index (BMI) ranging 18 28. All have been gym-goers least 1 year with no records chronic diseases, muscle, or bone surgeries. We encountered two main challenges while collecting our dataset. The first challenge was the dumbbell’s suitability, where found dumbbell weight (4.5 kg) provides best tradeoff longer recording sessions occurrence on exercises. second subjectivity RPE, average reported RPE measured heart rate converted RPE. observed data reduces biceps’ angular velocity; therefore, increases completion time later sets. extracted total 33 features dataset, which reduced 16 features. These are most overall representative correlated movement, yet they fatigue-specific utilized these in five machine learning models, Generalized Linear Models (GLM), Logistic Regression (LR), Random Forests (RF), Decision Trees (DT), Feedforward Neural Networks (FNN). using two-layer FNN achieves accuracy 98% 88% subject-specific cross-subject respectively. results presented work useful represent solid start moving into real-world application detecting level muscles sensors advise athletes take consideration avoid fatigue-induced injuries.