Segment-Based Unsupervised Learning Method in Sensor-Based Human Activity Recognition
Sensor-based human activity recognition (HAR) is a task to recognize human activities, and HAR has an important role in analyzing human behavior such as in the healthcare field. HAR is typically implemented using traditional machine learning methods. In contrast to traditional machine learning metho...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-10-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/23/20/8449 |
| _version_ | 1797572314876870656 |
|---|---|
| author | Koki Takenaka Kei Kondo Tatsuhito Hasegawa |
| author_facet | Koki Takenaka Kei Kondo Tatsuhito Hasegawa |
| author_sort | Koki Takenaka |
| collection | DOAJ |
| description | Sensor-based human activity recognition (HAR) is a task to recognize human activities, and HAR has an important role in analyzing human behavior such as in the healthcare field. HAR is typically implemented using traditional machine learning methods. In contrast to traditional machine learning methods, deep learning models can be trained end-to-end with automatic feature extraction from raw sensor data. Therefore, deep learning models can adapt to various situations. However, deep learning models require substantial amounts of training data, and annotating activity labels to construct a training dataset is cost-intensive due to the need for human labor. In this study, we focused on the continuity of activities and propose a segment-based unsupervised deep learning method for HAR using accelerometer sensor data. We define segment data as sensor data measured at one time, and this includes only a single activity. To collect the segment data, we propose a measurement method where the users only need to annotate the starting, changing, and ending points of their activity rather than the activity label. We developed a new segment-based SimCLR, which uses pairs of segment data, and propose a method that combines segment-based SimCLR with SDFD. We investigated the effectiveness of feature representations obtained by training the linear layer with fixed weights obtained by unsupervised learning methods. As a result, we demonstrated that the proposed combined method acquires generalized feature representations. The results of transfer learning on different datasets suggest that the proposed method is robust to the sampling frequency of the sensor data, although it requires more training data than other methods. |
| first_indexed | 2024-03-10T20:54:30Z |
| format | Article |
| id | doaj.art-2fab754e195a4ffaa4551f20ede0f4e5 |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-10T20:54:30Z |
| publishDate | 2023-10-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-2fab754e195a4ffaa4551f20ede0f4e52023-11-19T18:03:05ZengMDPI AGSensors1424-82202023-10-012320844910.3390/s23208449Segment-Based Unsupervised Learning Method in Sensor-Based Human Activity RecognitionKoki Takenaka0Kei Kondo1Tatsuhito Hasegawa2Graduate School of Engineering, University of Fukui, Fukui 910-8507, JapanGraduate School of Engineering, University of Fukui, Fukui 910-8507, JapanGraduate School of Engineering, University of Fukui, Fukui 910-8507, JapanSensor-based human activity recognition (HAR) is a task to recognize human activities, and HAR has an important role in analyzing human behavior such as in the healthcare field. HAR is typically implemented using traditional machine learning methods. In contrast to traditional machine learning methods, deep learning models can be trained end-to-end with automatic feature extraction from raw sensor data. Therefore, deep learning models can adapt to various situations. However, deep learning models require substantial amounts of training data, and annotating activity labels to construct a training dataset is cost-intensive due to the need for human labor. In this study, we focused on the continuity of activities and propose a segment-based unsupervised deep learning method for HAR using accelerometer sensor data. We define segment data as sensor data measured at one time, and this includes only a single activity. To collect the segment data, we propose a measurement method where the users only need to annotate the starting, changing, and ending points of their activity rather than the activity label. We developed a new segment-based SimCLR, which uses pairs of segment data, and propose a method that combines segment-based SimCLR with SDFD. We investigated the effectiveness of feature representations obtained by training the linear layer with fixed weights obtained by unsupervised learning methods. As a result, we demonstrated that the proposed combined method acquires generalized feature representations. The results of transfer learning on different datasets suggest that the proposed method is robust to the sampling frequency of the sensor data, although it requires more training data than other methods.https://www.mdpi.com/1424-8220/23/20/8449human activity recognitionunsupervised representation learningaccelerometer sensor datasegment data |
| spellingShingle | Koki Takenaka Kei Kondo Tatsuhito Hasegawa Segment-Based Unsupervised Learning Method in Sensor-Based Human Activity Recognition Sensors human activity recognition unsupervised representation learning accelerometer sensor data segment data |
| title | Segment-Based Unsupervised Learning Method in Sensor-Based Human Activity Recognition |
| title_full | Segment-Based Unsupervised Learning Method in Sensor-Based Human Activity Recognition |
| title_fullStr | Segment-Based Unsupervised Learning Method in Sensor-Based Human Activity Recognition |
| title_full_unstemmed | Segment-Based Unsupervised Learning Method in Sensor-Based Human Activity Recognition |
| title_short | Segment-Based Unsupervised Learning Method in Sensor-Based Human Activity Recognition |
| title_sort | segment based unsupervised learning method in sensor based human activity recognition |
| topic | human activity recognition unsupervised representation learning accelerometer sensor data segment data |
| url | https://www.mdpi.com/1424-8220/23/20/8449 |
| work_keys_str_mv | AT kokitakenaka segmentbasedunsupervisedlearningmethodinsensorbasedhumanactivityrecognition AT keikondo segmentbasedunsupervisedlearningmethodinsensorbasedhumanactivityrecognition AT tatsuhitohasegawa segmentbasedunsupervisedlearningmethodinsensorbasedhumanactivityrecognition |