This paper presents a novel method for fast classification of surface electromyography(sEMG) signals, using a simple model of attention. The brain transmits electrical signals throughout the body to contract and relax muscles. sEMG measures these signals by recording muscle activity from the surface above the muscle on the skin. By classifying these signals with low latency, they can be used to control a prosthetic limb using an amputee's brain power. On a difficult, industry benchmark sEMG dataset, the proposed attentional architecture yields excellent results, classifying 36 more gestures (53 in total) with about 20% higher accuracy (87% overall) than the current standards in the field. These results have direct and immediate application in the fields of robotics, myoelectric control, and prosthetics.
Josephs, John D. Jr; Drake, Carson; Cobb, Che; and Santerre, John
"sEMG Gesture Recognition With a Simple Model of Attention,"
SMU Data Science Review: Vol. 3:
1, Article 9.
Available at: https://scholar.smu.edu/datasciencereview/vol3/iss1/9
Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial 4.0 License