Significant Features for Human Activity Recognition Using Tri-Axial Accelerometers

Abstract

Activity recognition using wearable sensors has become essential for a variety of applications. Tri-axial accelerometers are the most widely used sensor for activity recognition. Although various features have been used to capture patterns and classify the accelerometer signals to recognise activities, there is no consensus on the best features to choose. Reducing the number of features can reduce the computational cost and complexity and enhance the performance of the classifiers. This paper identifies the signal features that have significant discriminative power between different human activities. It also investigates the effect of sensor placement location, the sampling frequency, and activity complexity on the selected features. A comprehensive list of 193 signal features has been extracted from accelerometer signals of four publicly available datasets, including features that have never been used before for activity recognition. Feature significance was measured using the Joint Mutual Information Maximisation (JMIM) method. Common significant features among all the datasets were identified. The results show that the sensor placement location does not significantly affect recognition performance, nor does it affect the significant sub-set of features. The results also showed that with high sampling frequency, features related to signal repeatability and regularity show high discriminative power.

Accuracy and Reliability of a Suite of Digital Measures of Walking Generated Using a Wrist-worn Sensor: Performance Characterization Study

“…Prior studies have used co-located investigational and reference sensors (e.g., two wrist-worn devices). Due to the known potential errors associated with body placement when capturing walking-related data [36][37][38][39], co-location could be vulnerable to bias towards overestimating performance, which our approach seeks to mitigate. Further, most studies have had a narrow focus on step counts [10][11][12][13][14][15][16], mostly in controlled laboratory environments and/or for limited time periods (e.g., single day in real-world setting).…”