Details
| Originalsprache | Englisch |
|---|---|
| Titel des Sammelwerks | Image Analysis and Recognition |
| Untertitel | 11th International Conference, ICIAR 2014, Proceedings |
| Herausgeber (Verlag) | Springer Verlag |
| Seiten | 48-57 |
| Seitenumfang | 10 |
| ISBN (elektronisch) | 9783319117546 |
| Publikationsstatus | Veröffentlicht - 10 Okt. 2014 |
| Veranstaltung | 11th International Conference on Image Analysis and Recognition, ICIAR 2014 - Vilamoura, Portugal Dauer: 22 Okt. 2014 → 24 Okt. 2014 |
Publikationsreihe
| Name | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
|---|---|
| Band | 8815 |
| ISSN (Print) | 0302-9743 |
| ISSN (elektronisch) | 1611-3349 |
Abstract
In this paper, a framework to recognize human actions from acceleration data is proposed. An important step for an accurate recognition is the preprocessing of input data and the following classification by the machine learning algorithm. In this paper, we suggest to combine Dynamic Time Warping (DTW) with Random Forest. The intention of using DTW is to pre-process the data to eliminate outliers and to align the time series. Many applications require more than one inertial sensor for an accurate prediction of actions. In this paper, nine inertial sensors are deployed to ensure an accurate recognition of actions. Further, sensor fusion approaches are introduced and the most promising strategy is shown. The proposed framework is evaluated on a self-recorded dataset consisting of six human actions. Each action was performed three times by 20 subjects. The dataset is publicly available for download.
ASJC Scopus Sachgebiete
- Mathematik (insg.)
- Theoretische Informatik
- Informatik (insg.)
Zitieren
- Standard
- Harvard
- Apa
- Vancouver
- BibTex
- RIS
Image Analysis and Recognition: 11th International Conference, ICIAR 2014, Proceedings. Springer Verlag, 2014. S. 48-57 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Band 8815).
Publikation: Beitrag in Buch/Bericht/Sammelwerk/Konferenzband › Aufsatz in Konferenzband › Forschung › Peer-Review
}
TY - GEN
T1 - Multi-sensor acceleration-based action recognition
AU - Baumann, Florian
AU - Schulz, Irina
AU - Rosenhahn, Bodo
N1 - Funding information: This work has been partially funded by the ERC within the starting grant Dynamic MinVIP.
PY - 2014/10/10
Y1 - 2014/10/10
N2 - In this paper, a framework to recognize human actions from acceleration data is proposed. An important step for an accurate recognition is the preprocessing of input data and the following classification by the machine learning algorithm. In this paper, we suggest to combine Dynamic Time Warping (DTW) with Random Forest. The intention of using DTW is to pre-process the data to eliminate outliers and to align the time series. Many applications require more than one inertial sensor for an accurate prediction of actions. In this paper, nine inertial sensors are deployed to ensure an accurate recognition of actions. Further, sensor fusion approaches are introduced and the most promising strategy is shown. The proposed framework is evaluated on a self-recorded dataset consisting of six human actions. Each action was performed three times by 20 subjects. The dataset is publicly available for download.
AB - In this paper, a framework to recognize human actions from acceleration data is proposed. An important step for an accurate recognition is the preprocessing of input data and the following classification by the machine learning algorithm. In this paper, we suggest to combine Dynamic Time Warping (DTW) with Random Forest. The intention of using DTW is to pre-process the data to eliminate outliers and to align the time series. Many applications require more than one inertial sensor for an accurate prediction of actions. In this paper, nine inertial sensors are deployed to ensure an accurate recognition of actions. Further, sensor fusion approaches are introduced and the most promising strategy is shown. The proposed framework is evaluated on a self-recorded dataset consisting of six human actions. Each action was performed three times by 20 subjects. The dataset is publicly available for download.
UR - http://www.scopus.com/inward/record.url?scp=84908663330&partnerID=8YFLogxK
U2 - 10.1007/978-3-319-11755-3_6
DO - 10.1007/978-3-319-11755-3_6
M3 - Conference contribution
AN - SCOPUS:84908663330
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 48
EP - 57
BT - Image Analysis and Recognition
PB - Springer Verlag
T2 - 11th International Conference on Image Analysis and Recognition, ICIAR 2014
Y2 - 22 October 2014 through 24 October 2014
ER -