WEKO3
アイテム
環境ビーコンを用いた屋内ローカライゼーションのためのリラベリングに基づくデータ補強
https://doi.org/10.18997/0002000946
https://doi.org/10.18997/000200094672154a93-3742-4fc5-9a1a-cd4d6363a7f5
| 名前 / ファイル | ライセンス | アクション |
|---|---|---|
sei_k_488.pdf (12.3 MB)
|
|
| アイテムタイプ | 学位論文 = Thesis or Dissertation(1) | |||||||
|---|---|---|---|---|---|---|---|---|
| 公開日 | 2024-09-03 | |||||||
| 資源タイプ | ||||||||
| 資源タイプ識別子 | http://purl.org/coar/resource_type/c_db06 | |||||||
| 資源タイプ | doctoral thesis | |||||||
| タイトル | ||||||||
| タイトル | Relabeling-based Data Augmentation for Indoor Localization with Environmental Beacons | |||||||
| 言語 | en | |||||||
| タイトル | ||||||||
| タイトル | 環境ビーコンを用いた屋内ローカライゼーションのためのリラベリングに基づくデータ補強 | |||||||
| 言語 | ja | |||||||
| 言語 | ||||||||
| 言語 | eng | |||||||
| 著者 |
Garcia, Christina
× Garcia, Christina
|
|||||||
| 抄録 | ||||||||
| 内容記述タイプ | Abstract | |||||||
| 内容記述 | In this thesis, we propose a new data augmentation method to improve real-life indoor localization in nursing care facilities using Bluetooth Low Energy (BLE) beacons for monitoring of patient and staff assistance, thereby aiding in workload balance. As more nursing homes are adopting Indoor Positioning Systems (IPS) with Internet of Things (IoT), this research aims to leverage sensor data collected in real-world environment to improve indoor localization. We introduce a novel relabeling approach to address the challenge of limited on-site data in beacon-based indoor localization. By analyzing signal patterns in different rooms, we successfully implement a relabeling strategy, utilizing Received Signal Strength Indicator (RSSI) values from one location as a proxy in another location with less samples. This method effectively expands the training set, enhancing model accuracy, as demonstrated in a nursing care facility where beacon devices and mobile applications were employed for data collection. Furthermore, the proposed the relabeling approach is tailored for stationary BLE beacons as positioned in indoor localization. This technique aims to resolve data imbalance by reusing RSS data from different sensors, effectively augmenting the dataset for machine learning applications. By measuring signal patterns using standard deviation and Kullback-Leibler divergence between minority and majority classes, we identified matching beacons to relabel. By matching beacons between classes, two variations of relabeling are implemented, specifically full and partial matching. A Random Forest model is utilized for location recognition, and performance is compared using the weighted F1-score to account for class imbalance. Our real-world dataset, collected over five days in a nursing facility, showcases the efficacy of this method, with a notable improvement in overall weighted F1-score compared to traditional augmentation techniques like Random Sampling, Synthetic Minority Oversampling Technique (SMOTE) and Adaptive Synthetic Sampling (ADASYN). Lastly, we delve into the domain of clinical pathways, highlighting the significance of human activity recognition (HAR) methods in analyzing movements affected by various medical conditions and professional activities, as outlined in clinical pathway documents. This understanding aids in refining decision-making processes and optimizing clinical pathways, with a focus on patient-centric studies and the monitoring of clinical activities. We underscore the potential of indoor localization in healthcare following the opportunities for activity recognition in clinical path to support management and asset tracking, patient and staff monitoring, navigation, and infection control. Leveraging existing sensor data and expanding the use of HAR, open avenues for more precise and efficient indoor localization solutions can be addressed. For future work, applying relabeling to other use-case scenario using other sensors aside from beacons is significant to understand signal patterns for matching. Moreover, relabeling approach can be further developed to cover locations in common area not exhibiting similar geometry and with unequal dimension such as the cafeteria and nurse station where nursing activities are performed longer. Expanding the analysis of signal patterns by incorporating additional statistical techniques is suggested to represent signal pattern feature of different sensors aside from BLE beacons. Hybrid augmentation approach by combining the proposed augmentation method with SMOTE or ADASYN is suggested to resolve limitations of respective augmentation techniques for other applications The integration of data augmentation methods that leverage sensor data onsite presents a promising avenue for enhancing indoor localization systems within nursing care facilities. This approach significantly contributes to the improvement of patient care and the effective management of workload among healthcare staff. |
|||||||
| 目次 | ||||||||
| 内容記述タイプ | TableOfContents | |||||||
| 内容記述 | 第1章 Introduction| 第2章 System Requirements for Indoor Positioning| 第3章 Signal Patterns and Relabeling Method| 第4章 Improving Relabeling with Full and Partial Matching| 第5章 Indoor Positioning with the Potential of Activity Recognition in Clinical Pathways| 第6章 Discussion and Future Work| 第7章 Conclusion | |||||||
| 備考 | ||||||||
| 内容記述タイプ | Other | |||||||
| 内容記述 | 九州⼯業⼤学博⼠学位論⽂ 学位記番号：生工博甲第488号 学位授与年⽉⽇: 令和6年3⽉25⽇ | |||||||
| 学位授与番号 | ||||||||
| 学位授与番号 | 甲第488号 | |||||||
| 学位名 | ||||||||
| 学位名 | 博士（工学） | |||||||
| 学位授与年月日 | ||||||||
| 学位授与年月日 | 2024-03-25 | |||||||
| 学位授与機関 | ||||||||
| 学位授与機関識別子Scheme | kakenhi | |||||||
| 学位授与機関識別子 | 17104 | |||||||
| 学位授与機関名 | 九州工業大学 | |||||||
| 言語 | ja | |||||||
| 学位授与年度 | ||||||||
| 内容記述タイプ | Other | |||||||
| 内容記述 | 令和5年度 | |||||||
| 出版タイプ | ||||||||
| 出版タイプ | VoR | |||||||
| 出版タイプResource | http://purl.org/coar/version/c_970fb48d4fbd8a85 | |||||||
| アクセス権 | ||||||||
| アクセス権 | open access | |||||||
| アクセス権URI | http://purl.org/coar/access_right/c_abf2 | |||||||
| ID登録 | ||||||||
| ID登録 | 10.18997/0002000946 | |||||||
| ID登録タイプ | JaLC | |||||||
