Name
Abstract | ||
|---|---|---|
In recent years, human motion recognition based on smartphones has gotten increasing attention in many fields such as mobile health, health tracking and pervasive computing. However, motion recognition performance can be easily affected by variation of phone orientations and positions. Different users have influence on recognition accuracy as well. Most of existing work focuses on one or two respects of above problems, or train different models for different phone positions and orientations. In this paper, we propose a generic framework for human motion recognition based on smartphones, which can effectively discriminate six daily motions regardless of device positions and orientations. We select a set of more robust and effective features to solve performance degradation problem caused by different phone positions, phone orientations and users. In experiments, we access our method using the dataset collected by three volunteers on Android smartphones. The experimental results show that the proposed feature extraction algorithm is better than most existing algorithms. |
Year | DOI | Venue |
|---|---|---|
2015 | 10.1109/IIKI.2015.71 | IIKI |
Keywords | Field | DocType |
Motion recognition, Smartphones, Feature extraction | Computer vision,Android (operating system),Accelerometer,Motion recognition,Computer science,Degradation Problem,Feature extraction,Human motion,Phone,Artificial intelligence,Ubiquitous computing | Conference |
Citations | PageRank | References |
0 | 0.34 | 10 |
Authors | ||
5 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Zhuorong Li | 1 | 9 | 1.57 |
| Xiaobo Xie | 2 | 25 | 6.81 |
| Xi Zhou | 3 | 48 | 10.28 |
| Junqi Guo | 4 | 61 | 15.07 |
| Rongfang Bie | 5 | 547 | 68.23 |