[B-19] Healthcare and Medical Information Communication Technology

11:15 AM - 11:30 AM JST(2:15 AM - 2:30 AM UTC)Candidate for Academic Incentive Award

[B-19-29] Development of an Interpretable Residual Compensation Model for Force Estimation from Surface Electromyography

◎Daiki Sodenaga Sodenaga¹, Seiichiro Katsura¹ (1. Keio University)

Keywords:

Biological signal,Sensing,Machine learning

This study aims to achieve force sensing without placing force sensors at the contact point by estimating human force using surface electromyography (sEMG), which is generated during muscle contraction and can be non-invasively measured from the skin surface. This paper proposes a model for estimating force from sEMG. Conventional model-based methods require retraining of model parameters due to fluctuations in the relationship between electromyographic signals and force between subjects and between experiments. This paper describes a calibration-free model-based force estimation method that expresses the residuals resulting from these fluctuations as an interpretable model.