Towards Physical Activity Recognition Using Smartphone Sensors

Abstract

In recent years, the use of a smartphone accelerometer in physical activity recognition has been well studied. However, the role of a gyroscope and a magnetometer is yet to be explored, both when used alone as well as in combination with an accelerometer. For this purpose, we investigate the role of these three smartphone sensors in activity recognition. We evaluate their roles on four body positions using seven classifiers while recognizing six physical activities. We show that in general an accelerometer and a gyroscope complement each other, thereby making the recognition process more reliable. Moreover, in most cases, a gyroscope does not only improve the recognition accuracy in combination with an accelerometer, but it also achieves a reasonable performance when used alone. The results for a magnetometer are not encouraging because it causes over-fitting in training classifiers due to its dependence on directions. Based on our evaluations, we show that it is difficult to make an exact general statement about which sensor performs better than the others in all situations because their recognition performance depends on the smartphone's position, the selected classifier, and the activity being recognized. However, statements about their roles in specific situations can be made. We report our observations and results in detail in this paper, while our data-set and datacollection app is publicly available, thereby making our experiments reproducible.

Sensor-based adaptive activity recognition with dynamically available sensors

“…An exception is found in the dataset SD-POCKET, in which the noBelief achieves the f1-score as high as 0.93, as gyroscope performs better than accelerometer in pocket position, confirmed by [21]. Therefore, initial model with gyroscope is able to correctly recognize most of the activities with high confidence, and provides true labels for the retraining with the combination of accelerometer and gyroscope data, hence the resulting model can then significantly improve the recognition performance.…”

“…As the examples classified with high confidence usually tend to be the correct classification, we do not update the posterior distribution for those high-confidence examples during the iterative process of belief propagation, so that their beliefs can be propagated to the uncertain examples. To demonstrate the effectiveness of belief propagation in smoothing the outliers, we perform physical activity recognition on smart phone sensor data from [21]. The mobile phone is fixed on the belt when the subject performing the activities, inertial data from accelerometer and gyroscope are collected, which is known to be effective for physical activity recognition.…”

“…Smartphone dataset (SD) [21]: Activity data was collected from accelerometer, gyroscope and magnetometer on an Android device worn in different body position (arm, belt, waist and pocket), when the subject performs standing, walking, upstairs, sitting, running and downstairs. The sample rate is set to be 50Hz.…”

“…Sensors activity dataset (SAD) [21]: Sensor data was collected when the 10 volunteers perform standing, walking, upstairs, sitting, downstairs, jogging and biking. We extract the same features as the first dataset.…”

Mobile activity recognition and fall detection system for elderly people using Ameva algorithm

“…• Shoaib [42]: this dataset contains smartphone sensor data for six physical activities collected using four participants. It was useful because the data was collected from four smartphones on four body positions, allowing for comparison to approaches using only one smartphone.…”