Domain-generalization human activity recognition model based on CSI instance normalization
To achieve Wi-Fi cross-domain human activity perception that was not dependent on target domain data, a domain-generalization human activity recognition model based on CSI instance normalization called INDG-Fi was proposed. The instance normalization standardization was utilized to remove domain inf...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | zho |
| Published: |
Editorial Department of Journal on Communications
2024-06-01
|
| Series: | Tongxin xuebao |
| Subjects: | |
| Online Access: | http://www.joconline.com.cn/zh/article/doi/10.11959/j.issn.1000-436x.2024110/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841539242169204736 |
|---|---|
| author | WANG Yang XU Jiawei WANG Ao XIA Huijuan ZHAO Chuanxin JI Yimu |
| author_facet | WANG Yang XU Jiawei WANG Ao XIA Huijuan ZHAO Chuanxin JI Yimu |
| author_sort | WANG Yang |
| collection | DOAJ |
| description | To achieve Wi-Fi cross-domain human activity perception that was not dependent on target domain data, a domain-generalization human activity recognition model based on CSI instance normalization called INDG-Fi was proposed. The instance normalization standardization was utilized to remove domain information from the representation of CSI features by INDG-Fi. Then action classifiers and domain classifiers were constructed for shared feature extraction. By employing activity bias learning and adversarial domain learning, the model biased the features extracted from the encoding layer towards signal variations caused by human actions while moving away from domain signals. To enhance the model’s focus on subcarrier signals that were more significantly influenced by human actions, a subcarrier attention module was incorporated into the encoding layer. The implemented results demonstrate that the proposed INDG-Fi achieves perceptual accuracies of 97.99% and 92.73% for unseen users and locations, respectively, thus enabling robust cross-domain perception. |
| format | Article |
| id | doaj-art-eec756f0a8ec40e1bc86b9dc5781db82 |
| institution | Kabale University |
| issn | 1000-436X |
| language | zho |
| publishDate | 2024-06-01 |
| publisher | Editorial Department of Journal on Communications |
| record_format | Article |
| series | Tongxin xuebao |
| spelling | doaj-art-eec756f0a8ec40e1bc86b9dc5781db822025-01-14T07:24:34ZzhoEditorial Department of Journal on CommunicationsTongxin xuebao1000-436X2024-06-014519620963977505Domain-generalization human activity recognition model based on CSI instance normalizationWANG YangXU JiaweiWANG AoXIA HuijuanZHAO ChuanxinJI YimuTo achieve Wi-Fi cross-domain human activity perception that was not dependent on target domain data, a domain-generalization human activity recognition model based on CSI instance normalization called INDG-Fi was proposed. The instance normalization standardization was utilized to remove domain information from the representation of CSI features by INDG-Fi. Then action classifiers and domain classifiers were constructed for shared feature extraction. By employing activity bias learning and adversarial domain learning, the model biased the features extracted from the encoding layer towards signal variations caused by human actions while moving away from domain signals. To enhance the model’s focus on subcarrier signals that were more significantly influenced by human actions, a subcarrier attention module was incorporated into the encoding layer. The implemented results demonstrate that the proposed INDG-Fi achieves perceptual accuracies of 97.99% and 92.73% for unseen users and locations, respectively, thus enabling robust cross-domain perception.http://www.joconline.com.cn/zh/article/doi/10.11959/j.issn.1000-436x.2024110/channel state informationwireless sensinghuman activity recognitiondomain-generalization |
| spellingShingle | WANG Yang XU Jiawei WANG Ao XIA Huijuan ZHAO Chuanxin JI Yimu Domain-generalization human activity recognition model based on CSI instance normalization Tongxin xuebao channel state information wireless sensing human activity recognition domain-generalization |
| title | Domain-generalization human activity recognition model based on CSI instance normalization |
| title_full | Domain-generalization human activity recognition model based on CSI instance normalization |
| title_fullStr | Domain-generalization human activity recognition model based on CSI instance normalization |
| title_full_unstemmed | Domain-generalization human activity recognition model based on CSI instance normalization |
| title_short | Domain-generalization human activity recognition model based on CSI instance normalization |
| title_sort | domain generalization human activity recognition model based on csi instance normalization |
| topic | channel state information wireless sensing human activity recognition domain-generalization |
| url | http://www.joconline.com.cn/zh/article/doi/10.11959/j.issn.1000-436x.2024110/ |
| work_keys_str_mv | AT wangyang domaingeneralizationhumanactivityrecognitionmodelbasedoncsiinstancenormalization AT xujiawei domaingeneralizationhumanactivityrecognitionmodelbasedoncsiinstancenormalization AT wangao domaingeneralizationhumanactivityrecognitionmodelbasedoncsiinstancenormalization AT xiahuijuan domaingeneralizationhumanactivityrecognitionmodelbasedoncsiinstancenormalization AT zhaochuanxin domaingeneralizationhumanactivityrecognitionmodelbasedoncsiinstancenormalization AT jiyimu domaingeneralizationhumanactivityrecognitionmodelbasedoncsiinstancenormalization |