Toward Self-Calibrating Plug-and-Play Myoelectric Control
DOI:
https://doi.org/10.57922/mec.2480Abstract
Myoelectric control enables users to interact with diverse devices. However, electromyographic (EMG) signals change over time due to diverse factors, e.g., user behaviour variation, and other. These variations lead to a substantial reduction in performance of machine learning-based myoelectric control model, which in turn necessitate frequent re-calibration. In this paper, we report the results of our “self-calibrating” and “plug-and-play” random forest model. We pre-train the model and then calibrate it on new participants via one-shot calibration. The model then calibrate itself autonomously . We validated this model on 18 testing participants. Work is on-going to expand our database and study the effectiveness of the approach with people with limb difference.
Downloads
Published
2024-08-15
How to Cite
[1]
X. Jiang, C. Ma, and K. Nazarpour, “Toward Self-Calibrating Plug-and-Play Myoelectric Control”, MEC Symposium, Aug. 2024.
Conference Proceedings Volume
Section
Myoelectric Control Algorithms
