2020/07/29 by Cheikh Latyr Fall, Fall, Cheikh Latyr, Ulysse Côté Allard +12
Computer Science · Engineering · Medicine · #Context-Aware Activity Recognition Systems #FOS: Computer and information sciences #Gaze Tracking and Assistive Technology #Human-Computer Interaction (cs.HC) #Muscle activation and electromyography studies #Robotics (cs.RO) #Stroke Rehabilitation and Recovery
paper · pdf · doi:10.48550/arxiv.2007.15032
openalex publication_date 2020/07/29 · openalex created_date 2022/07/26 · openalex updated_date 2026/07/28
This paper presents a control interface to translate the residual body\nmotions of individuals living with severe disabilities, into control commands\nfor body-machine interaction. A custom, wireless, wearable multi-sensor network\nis used to collect motion data from multiple points on the body in real-time.\nThe solution proposed successfully leverage electromyography gesture\nrecognition techniques for the recognition of inertial measurement units-based\ncommands (IMU), without the need for cumbersome and noisy surface electrodes.\nMotion pattern recognition is performed using a computationally inexpensive\nclassifier (Linear Discriminant Analysis) so that the solution can be deployed\nonto lightweight embedded platforms. Five participants (three able-bodied and\ntwo living with upper-body disabilities) presenting different motion\nlimitations (e.g. spasms, reduced motion range) were recruited. They were asked\nto perform up to 9 different motion classes, including head, shoulder, finger,\nand foot motions, with respect to their residual functional capacities. The\nmeasured prediction performances show an average accuracy of 99.96% for\nable-bodied individuals and 91.66% for participants with upper-body\ndisabilities. The recorded dataset has also been made available online to the\nresearch community. Proof of concept for the real-time use of the system is\ngiven through an assembly task replicating activities of daily living using the\nJACO arm from Kinova Robotics.\n