Iterative Self-Training Based Domain Adaptation for Cross-User sEMG Gesture Recognition
Surface electromyography (sEMG) based gesture recognition has received broad attention and application in rehabilitation areas for its direct and fine-grained sensing ability. sEMG signals exhibit strong user dependence properties among users with different physiology, causing the inapplicability of...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2023-01-01
|
| Series: | IEEE Transactions on Neural Systems and Rehabilitation Engineering |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10175382/ |
| _version_ | 1797771683860316160 |
|---|---|
| author | Kang Wang Yiqiang Chen Yingwei Zhang Xiaodong Yang Chunyu Hu |
| author_facet | Kang Wang Yiqiang Chen Yingwei Zhang Xiaodong Yang Chunyu Hu |
| author_sort | Kang Wang |
| collection | DOAJ |
| description | Surface electromyography (sEMG) based gesture recognition has received broad attention and application in rehabilitation areas for its direct and fine-grained sensing ability. sEMG signals exhibit strong user dependence properties among users with different physiology, causing the inapplicability of the recognition model on new users. Domain adaptation is the most representative method to reduce the user gap with feature decoupling to acquire motion-related features. However, the existing domain adaptation method shows awful decoupling results when handling complex time-series physiological signals. Therefore, this paper proposes an Iterative Self-Training based Domain Adaptation method (STDA) to supervise the feature decoupling process with the pseudo-label generated by self-training and to explore cross-user sEMG gesture recognition. STDA mainly consists of two parts, discrepancy-based domain adaptation (DDA) and pseudo-label iterative update (PIU). DDA aligns existing users’ data and new users’ unlabeled data with a Gaussian kernel-based distance constraint. PIU Iteratively continuously updates pseudo-labels to generate more accurate labelled data on new users with category balance. Detailed experiments are performed on publicly available benchmark datasets, including the NinaPro dataset (DB-1 and DB-5) and the CapgMyo dataset (DB-a, DB-b, and DB-c). Experimental results show that the proposed method achieves significant performance improvement compared with existing sEMG gesture recognition and domain adaption methods. |
| first_indexed | 2024-03-12T21:41:07Z |
| format | Article |
| id | doaj.art-1ceffdaa700343df9bcf5d4f6376992c |
| institution | Directory Open Access Journal |
| issn | 1558-0210 |
| language | English |
| last_indexed | 2024-03-12T21:41:07Z |
| publishDate | 2023-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Transactions on Neural Systems and Rehabilitation Engineering |
| spelling | doaj.art-1ceffdaa700343df9bcf5d4f6376992c2023-07-26T23:00:02ZengIEEEIEEE Transactions on Neural Systems and Rehabilitation Engineering1558-02102023-01-01312974298710.1109/TNSRE.2023.329333410175382Iterative Self-Training Based Domain Adaptation for Cross-User sEMG Gesture RecognitionKang Wang0https://orcid.org/0009-0001-2976-8764Yiqiang Chen1https://orcid.org/0000-0002-8407-0780Yingwei Zhang2https://orcid.org/0000-0002-6582-1745Xiaodong Yang3https://orcid.org/0000-0002-1842-5475Chunyu Hu4School of Computer Science and Technology, Qilu University of Technology (Shandong Academy of Sciences), Jinan, ChinaShangdong Academy of Intelligent Computing Technology, Jinan, ChinaShangdong Academy of Intelligent Computing Technology, Jinan, ChinaShangdong Academy of Intelligent Computing Technology, Jinan, ChinaSchool of Computer Science and Technology, Qilu University of Technology (Shandong Academy of Sciences), Jinan, ChinaSurface electromyography (sEMG) based gesture recognition has received broad attention and application in rehabilitation areas for its direct and fine-grained sensing ability. sEMG signals exhibit strong user dependence properties among users with different physiology, causing the inapplicability of the recognition model on new users. Domain adaptation is the most representative method to reduce the user gap with feature decoupling to acquire motion-related features. However, the existing domain adaptation method shows awful decoupling results when handling complex time-series physiological signals. Therefore, this paper proposes an Iterative Self-Training based Domain Adaptation method (STDA) to supervise the feature decoupling process with the pseudo-label generated by self-training and to explore cross-user sEMG gesture recognition. STDA mainly consists of two parts, discrepancy-based domain adaptation (DDA) and pseudo-label iterative update (PIU). DDA aligns existing users’ data and new users’ unlabeled data with a Gaussian kernel-based distance constraint. PIU Iteratively continuously updates pseudo-labels to generate more accurate labelled data on new users with category balance. Detailed experiments are performed on publicly available benchmark datasets, including the NinaPro dataset (DB-1 and DB-5) and the CapgMyo dataset (DB-a, DB-b, and DB-c). Experimental results show that the proposed method achieves significant performance improvement compared with existing sEMG gesture recognition and domain adaption methods.https://ieeexplore.ieee.org/document/10175382/Surface electromyographygesture recognitioncross-userdomain adaptationsemi-supervised learning |
| spellingShingle | Kang Wang Yiqiang Chen Yingwei Zhang Xiaodong Yang Chunyu Hu Iterative Self-Training Based Domain Adaptation for Cross-User sEMG Gesture Recognition IEEE Transactions on Neural Systems and Rehabilitation Engineering Surface electromyography gesture recognition cross-user domain adaptation semi-supervised learning |
| title | Iterative Self-Training Based Domain Adaptation for Cross-User sEMG Gesture Recognition |
| title_full | Iterative Self-Training Based Domain Adaptation for Cross-User sEMG Gesture Recognition |
| title_fullStr | Iterative Self-Training Based Domain Adaptation for Cross-User sEMG Gesture Recognition |
| title_full_unstemmed | Iterative Self-Training Based Domain Adaptation for Cross-User sEMG Gesture Recognition |
| title_short | Iterative Self-Training Based Domain Adaptation for Cross-User sEMG Gesture Recognition |
| title_sort | iterative self training based domain adaptation for cross user semg gesture recognition |
| topic | Surface electromyography gesture recognition cross-user domain adaptation semi-supervised learning |
| url | https://ieeexplore.ieee.org/document/10175382/ |
| work_keys_str_mv | AT kangwang iterativeselftrainingbaseddomainadaptationforcrossusersemggesturerecognition AT yiqiangchen iterativeselftrainingbaseddomainadaptationforcrossusersemggesturerecognition AT yingweizhang iterativeselftrainingbaseddomainadaptationforcrossusersemggesturerecognition AT xiaodongyang iterativeselftrainingbaseddomainadaptationforcrossusersemggesturerecognition AT chunyuhu iterativeselftrainingbaseddomainadaptationforcrossusersemggesturerecognition |