| Title: | HARE : unifying the human activity recognition engineering workflow |
|---|
| Authors: | ID Konak, Orhan (Author)
ID Liebe, Lucas (Author)
ID Postnov, Kirill (Author)
ID Sauerwald, Franz (Author)
ID Gjoreski, Hristijan (Author)
ID Luštrek, Mitja, Institut Jožef Stefan (Author)
ID Arnrich, Bert (Author) |
|---|
| Files: |  URL - Source URL, visit https://www.mdpi.com/1424-8220/23/23/9571
 PDF - Presentation file, download (6,40 MB)
MD5: 77022A9114AE9FB2ACB53384AF973E4A
|
|---|
| Language: | English |
|---|
| Typology: | 1.01 - Original Scientific Article |
|---|
| Organization: |  IJS - Jožef Stefan Institute
|
|---|
| Abstract: | Sensor-based human activity recognition is becoming ever more prevalent. The increasing importance of distinguishing human movements, particularly in healthcare, coincides with the advent of increasingly compact sensors. A complex sequence of individual steps currently characterizes the activity recognition pipeline. It involves separate data collection, preparation, and processing steps, resulting in a heterogeneous and fragmented process. To address these challenges, we present a comprehensive framework, HARE, which seamlessly integrates all necessary steps. HARE offers synchronized data collection and labeling, integrated pose estimation for data anonymization, a multimodal classification approach, and a novel method for determining optimal sensor placement to enhance classification results. Additionally, our framework incorporates real-time activity recognition with on-device model adaptation capabilities. To validate the effectiveness of our framework, we conducted extensive evaluations using diverse datasets, including our own collected dataset focusing on nursing activities. Our results show that HARE’s multimodal and on-device trained model outperforms conventional single-modal and offline variants. Furthermore, our vision-based approach for optimal sensor placement yields comparable results to the trained model. Our work advances the field of sensor-based human activity recognition by introducing a comprehensive framework that streamlines data collection and classification while offering a novel method for determining optimal sensor placement. |
|---|
| Keywords: | human activity recognition, multimodal classification, privacy preservation, real-time classification, sensor placement |
|---|
| Publication status: | Published |
|---|
| Publication version: | Version of Record |
|---|
| Submitted for review: | 13.10.2023 |
|---|
| Article acceptance date: | 29.11.2023 |
|---|
| Publication date: | 02.12.2023 |
|---|
| Publisher: | MDPI |
|---|
| Year of publishing: | 2023 |
|---|
| Number of pages: | str. 1-23 |
|---|
| Numbering: | Vol. 23, iss. 23, [article no.] 9571 |
|---|
| Source: | Švica |
|---|
| PID: | 20.500.12556/DiRROS-17496  |
|---|
| UDC: | 681.5 |
|---|
| ISSN on article: | 1424-8220 |
|---|
| DOI: | 10.3390/s23239571 |
|---|
| COBISS.SI-ID: | 174721027 |
|---|
| Copyright: | © 2023 by the authors |
|---|
| Note: | Nasl. z nasl. zaslona;
Soavtor iz Slovenije: Mitja Luštrek;
Opis vira z dne 4. 12. 2023;
|
|---|
| Publication date in DiRROS: | 11.12.2023 |
|---|
| Views: | 666 |
|---|
| Downloads: | 265 |
|---|
| Metadata: |    |
|---|
|
:
|
Copy citation |
|---|
| | | | Share: |  |
|---|
Hover the mouse pointer over a document title to show the abstract or click
on the title to get all document metadata. |
