| Presentation | 2023-01-19 [Short Paper] A Study on Improving the Robustness of Wi-Fi Sensing against Environmental Change Sorachi Kato, Tomoki Murakami, Takuya Fujihashi, Takashi Watanabe, Shunsuke Saruwatari, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | There is research on invasive sensing of objects using Channel State Information (CSI), which represents the propagation state of Wi-Fi radio waves. Neural networks are used to analyze CSI fluctuations, and have been shown to achieve high sensing accuracy in the fields of location estimation and activity recognition. On the other hand, there is a problem that sensing accuracy is significantly degraded when a new CSI acquired at a different time and environment from the one used for training. Re-training or transfer learning is necessary to adapt the model to the new environment, but in some practical applications, only a small amount of data from the new environment can be obtained, or sometimes none at all, due to the highly specialized methods used to collect label data for training. In this paper, we propose a method for respiratory rate estimation that transforms CSI features in different environments into a single domain, thereby guaranteeing sensing accuracy without label data in those environments. The method has a domain adaptation module that transforms CSI features acquired in different environments into a single domain based on the idea of style transformation in the image processing field. The method also uses Template of CSI and respiration sensor data, which are pre-obtained to include information on respiration at various frequencies. By adapting Template CSI to the characteristics of CSI in different environments and combining it with Template respiration sensor data to train a convolutional neural network (CNN), it is possible to train models that adapt to different environments using only unlabeled CSI in different environments. This paper shows that the mean absolute percentage error of the proposed method is reduced by about 20% compared to the case without domain adaptation. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | IEEE 802.11 / Wireless Sensing / Domain Adaptation / Vital Sensing / Wi-Fi |
| Paper # | SeMI2022-82 |
| Date of Issue | 2023-01-12 (SeMI) |
| Conference Information | |
| Committee | SeMI |
|---|---|
| Conference Date | 2023/1/19(2days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | Naruto grand hotel |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | |
| Chair | Koji Yamamoto(Kyoto Univ.) |
| Vice Chair | Kazuya Monden(Hitachi) / Yasunori Owada(NICT) / Shunsuke Saruwatari(Osaka Univ.) |
| Secretary | Kazuya Monden(NTT DOCOMO) / Yasunori Owada(Tokyo Univ. of Agri. and Tech.) / Shunsuke Saruwatari(Osaka Univ.) |
| Assistant | Yuki Matsuda(NAIST) / Akihito Taya(Aoyama Gakuin Univ.) / Takeshi Hirai(Osaka Univ.) |
| Paper Information | |
| Registration To | Technical Committee on Sensor Network and Mobile Intelligence |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | [Short Paper] A Study on Improving the Robustness of Wi-Fi Sensing against Environmental Change |
| Sub Title (in English) | |
| Keyword(1) | IEEE 802.11 |
| Keyword(2) | Wireless Sensing |
| Keyword(3) | Domain Adaptation |
| Keyword(4) | Vital Sensing |
| Keyword(5) | Wi-Fi |
| 1st Author's Name | Sorachi Kato |
| 1st Author's Affiliation | Osaka University(OU) |
| 2nd Author's Name | Tomoki Murakami |
| 2nd Author's Affiliation | Access Network Service Systems Laboratories, Nippon Telegraph and Telephone Corporation(NTT) |
| 3rd Author's Name | Takuya Fujihashi |
| 3rd Author's Affiliation | Osaka University(OU) |
| 4th Author's Name | Takashi Watanabe |
| 4th Author's Affiliation | Osaka University(OU) |
| 5th Author's Name | Shunsuke Saruwatari |
| 5th Author's Affiliation | Osaka University(OU) |
| Date | 2023-01-19 |
| Paper # | SeMI2022-82 |
| Volume (vol) | vol.122 |
| Number (no) | SeMI-341 |
| Page | pp.pp.49-50(SeMI), |
| #Pages | 2 |
| Date of Issue | 2023-01-12 (SeMI) |