physical activity recognition
Recently Published Documents


TOTAL DOCUMENTS

101
(FIVE YEARS 31)

H-INDEX

19
(FIVE YEARS 4)

2021 ◽  
Vol 9 (4) ◽  
pp. 0-0

This article describes a new scheme for a physical activity recognition process based on carried smartphone embedded sensors, such as accelerometer and gyroscope. For this purpose, the WKNN-SVM algorithm has been proposed to predict physical activities such as Walking, Standing or Sitting. It combines Weighted K-Nearest Neighbours (WKNN) and Support Vector Machines (SVM). The signals generated from the sensors are processed and then reduced using the Kernel Discriminant Analysis (KDA) by selecting the best discriminating components of the data. We performed different tests on four public datasets where the participants performed different activities carrying a smartphone. We demonstrated through several experiments that KDA/WKNN-SVM algorithm can improve the overall recognition performances, and has a higher recognition rate than the baseline methods using the machine learning and deep learning algorithms.


Author(s):  
Anastasia Motrenko ◽  
Egor Simchuk ◽  
Renat Khairullin ◽  
Andrey Inyakin ◽  
Daniil Kashirin ◽  
...  

2021 ◽  
Vol 11 (6) ◽  
pp. 2633
Author(s):  
Nora Alhammad ◽  
Hmood AlDossari

Data segmentation is an essential process in activity recognition when using machine learning techniques. Previous studies on physical activity recognition have mostly relied on the sliding window approach for segmentation. However, choosing a fixed window size for multiple activities with different durations may affect recognition accuracy, especially when the activities belong to the same category (i.e., dynamic or static). This paper presents and verifies a new method for dynamic segmentation of physical activities performed during the rehabilitation of individuals with spinal cord injuries. To adaptively segment the raw data, signal characteristics are analyzed to determine the suitable type of boundaries. Then, the algorithm identifies the time boundaries to represent the start- and endpoints of each activity. To verify the method and build a predictive model, an experiment was conducted in which data were collected using a single wrist-worn accelerometer sensor. The experimental results were compared with the sliding window approach, indicating that the proposed method outperformed the sliding window approach in terms of overall accuracy, which exceeded 5%, as well as model robustness. The results also demonstrated efficient physical activity segmentation using the proposed method, resulting in high classification performance for all activities considered.


Author(s):  
Wenda Li ◽  
Mohammud J. Bocus ◽  
Chong Tang ◽  
Robert. J. Piechocki ◽  
Karl Woodbridge ◽  
...  

Sign in / Sign up

Export Citation Format

Share Document