Application of gesture recognition based on domain transfer learning and SimSiam network in education management
As one of the important components of augmented reality system, gesture recognition is widely used in virtual reality, intelligent interaction and human-machine interface. In education management, how to improve the level of teaching management and teaching quality, gesture recognition plays a very...
Saved in:
| Main Author: | |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Tamkang University Press
2025-03-01
|
| Series: | Journal of Applied Science and Engineering |
| Subjects: | |
| Online Access: | http://jase.tku.edu.tw/articles/jase-202510-28-10-0019 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849390164203274240 |
|---|---|
| author | Xiaoxu He |
| author_facet | Xiaoxu He |
| author_sort | Xiaoxu He |
| collection | DOAJ |
| description | As one of the important components of augmented reality system, gesture recognition is widely used in virtual reality, intelligent interaction and human-machine interface. In education management, how to improve the level of teaching management and teaching quality, gesture recognition plays a very important role. In view of the problems of negative transfer and poor model generalization in single-source domain transfer learning, which have great impact on gesture recognition, this paper proposes a new novel gesture recognition method based on domain transfer learning and SimSiam network. Firstly, the gesture data set is passed into the SimSiam self-supervised network for training. On this basis, the technique of domain specific feature alignment
and domain classifier alignment is adopted. This approach aims to enhance the model’s gesture recognition performance between different users, thus significantly improving the accuracy of cross-user gesture recognition systems. Experimental results show that this proposed method can effectively identify a variety of dynamic gestures with Angle deflection. Compared with the traditional displacement feature method, the average accuracy of the proposed method is increased by 4%, and it can effectively deal with the dynamic gesture recognition problem in the case of palm deflection. |
| format | Article |
| id | doaj-art-4fc1ff1a413145a6aab6fdedf275cd2e |
| institution | Kabale University |
| issn | 2708-9967 2708-9975 |
| language | English |
| publishDate | 2025-03-01 |
| publisher | Tamkang University Press |
| record_format | Article |
| series | Journal of Applied Science and Engineering |
| spelling | doaj-art-4fc1ff1a413145a6aab6fdedf275cd2e2025-08-20T03:41:46ZengTamkang University PressJournal of Applied Science and Engineering2708-99672708-99752025-03-0128102067207510.6180/jase.202510_28(10).0019Application of gesture recognition based on domain transfer learning and SimSiam network in education managementXiaoxu He0Shenyang City University, No. 2, Wutong Street., Sujiatun District., Shenyang City, 110100 ChinaAs one of the important components of augmented reality system, gesture recognition is widely used in virtual reality, intelligent interaction and human-machine interface. In education management, how to improve the level of teaching management and teaching quality, gesture recognition plays a very important role. In view of the problems of negative transfer and poor model generalization in single-source domain transfer learning, which have great impact on gesture recognition, this paper proposes a new novel gesture recognition method based on domain transfer learning and SimSiam network. Firstly, the gesture data set is passed into the SimSiam self-supervised network for training. On this basis, the technique of domain specific feature alignment and domain classifier alignment is adopted. This approach aims to enhance the model’s gesture recognition performance between different users, thus significantly improving the accuracy of cross-user gesture recognition systems. Experimental results show that this proposed method can effectively identify a variety of dynamic gestures with Angle deflection. Compared with the traditional displacement feature method, the average accuracy of the proposed method is increased by 4%, and it can effectively deal with the dynamic gesture recognition problem in the case of palm deflection.http://jase.tku.edu.tw/articles/jase-202510-28-10-0019education managementgesture recognitiondomain transfer learningsimsiam network |
| spellingShingle | Xiaoxu He Application of gesture recognition based on domain transfer learning and SimSiam network in education management Journal of Applied Science and Engineering education management gesture recognition domain transfer learning simsiam network |
| title | Application of gesture recognition based on domain transfer learning and SimSiam network in education management |
| title_full | Application of gesture recognition based on domain transfer learning and SimSiam network in education management |
| title_fullStr | Application of gesture recognition based on domain transfer learning and SimSiam network in education management |
| title_full_unstemmed | Application of gesture recognition based on domain transfer learning and SimSiam network in education management |
| title_short | Application of gesture recognition based on domain transfer learning and SimSiam network in education management |
| title_sort | application of gesture recognition based on domain transfer learning and simsiam network in education management |
| topic | education management gesture recognition domain transfer learning simsiam network |
| url | http://jase.tku.edu.tw/articles/jase-202510-28-10-0019 |
| work_keys_str_mv | AT xiaoxuhe applicationofgesturerecognitionbasedondomaintransferlearningandsimsiamnetworkineducationmanagement |