A Feature Extraction Pipeline for Enhancing Lightweight Neural Networks in sEMG-based Joint Torque Estimation

TL;DR

This paper presents a feature extraction pipeline for sEMG signals that enables lightweight neural networks to accurately predict joint torques in robot-assisted rehabilitation, especially useful with limited training data.

Contribution

The study introduces a novel feature extraction pipeline that allows simple neural networks like MLP to perform comparably to complex models in torque prediction from sEMG signals.

Findings

MLP achieved mean RMSE of 0.963, 1.403, and 1.434 N·m for different joints.

The feature pipeline enables simple models to match complex network performance.

Results are promising for applications with limited training data.

Abstract

Robot-assisted rehabilitation offers an effective approach, wherein exoskeletons adapt to users' needs and provide personalized assistance. However, to deliver such assistance, accurate prediction of the user's joint torques is essential. In this work, we propose a feature extraction pipeline using 8-channel surface electromyography (sEMG) signals to predict elbow and shoulder joint torques. For preliminary evaluation, this pipeline was integrated into two neural network models: the Multilayer Perceptron (MLP) and the Temporal Convolutional Network (TCN). Data were collected from a single subject performing elbow and shoulder movements under three load conditions (0 kg, 1.10 kg, and 1.85 kg) using three motion-capture cameras. Reference torques were estimated from center-of-mass kinematics under the assumption of static equilibrium. Our offline analyses showed that, with our feature extraction pipeline, MLP model achieved mean RMSE of 0.963 N m, 1.403 N m, and 1.434 N m (over five seeds) for elbow, front-shoulder, and side-shoulder joints, respectively, which were comparable to the TCN performance. These results demonstrate that the proposed feature extraction pipeline enables a simple MLP to achieve performance comparable to that of a network designed explicitly for temporal dependencies. This finding is particularly relevant for applications with limited training data, a common scenario patient care.