Abstract
The effort, focus, and time to collect data and train EMG pattern recognition systems is one of the largest barriers to their widespread adoption in commercial applications. In addition to multiple repetitions of motions, including exemplars of confounding factors during the training protocol has been shown to be critical for robust machine learning models. This added training burden is prohibitive for most regular use cases, so cross-user models have been proposed that could leverage inter-repetition variability supplied by other users. Existing cross-user models have not yet achieved performance levels sufficient for commercialization and require users to closely adhere to a training protocol that is impractical without expert guidance. In this work, we extend a previously reported adaptive domain adversarial neural network (ADANN) to a cross-subject framework that requires very little training data from the end-user. We compare its performance to single-repetition within-user training and the previous state-of-the-art cross-subject technique, canonical correlation analysis (CCA). ADANN significantly outperformed CCA for both intact-limb (86.8–96.2%) and amputee (64.1–84.2%) populations. Moreover, the ADANN adaptation computation time was substantially lower than the time otherwise devoted to conducting a full within-subject training protocol. This study shows that cross-user models, enabled by deep-learned adaptations, may be a viable option for improved generalized pattern recognition-based myoelectric control.
Highlights
Electromyography (EMG) pattern recognition has been pursued as a way to control prosthetic devices for decades; there is a growing interest in its use for a wider range of commercial applications (Jiang et al, 2018)
The one-way repeatedmeasures analysis of variance (RMANOVA) investigating the effect of population found a significant effect for all pipelines except time domain (TD), TDAR, TDPSD, LSF4, and LSF9 in the crosssubject evaluation framework (p = 0.191, 0.181, 0.497, 0.082, 0.07, respectively)
Post-hoc tests revealed adversarial neural network (ADANN) performed significantly better for intact-limb subjects than all other pipelines across evaluation frameworks (p < 0.05), except the within-subject framework TDAR, TDPSD, LSF4, LSF9, and convolutional neural networks (CNN) pipelines (p = 0.058, 0.537, 0.704, 0.952, 0.174, respectively)
Summary
Electromyography (EMG) pattern recognition has been pursued as a way to control prosthetic devices for decades; there is a growing interest in its use for a wider range of commercial applications (Jiang et al, 2018). When multiple electrodes are used, the decoded signals may be used to infer the gesture performed by the user with the help of a pattern recognition model (Oskoei and Hu, 2008) These systems require collecting EMG data from the end-user through a guided training protocol before the device is used to configure the pattern recognition model. Following this procedure, high gesture recognition accuracy can be obtained under controlled settings, as was demonstrated by Côté-Allard et al, who obtained 98.3% for a 7 class system (Côté-Allard et al, 2019). When systems are used in real-world conditions, Deep Cross-User Gesture Recognition performance tends to degrade substantially due to confounding factors, such as limb position, contraction intensity, and electrode shift (Scheme and Englehart, 2011; Campbell et al, 2020c; Phinyomark et al, 2020)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.