Daily Living Activity Recognition In-The-Wild: Modeling and Inferring Activity-Aware Human Contexts

Abstract

Advancement in smart sensing and computing technologies has provided a dynamic opportunity to develop intelligent systems for human activity monitoring and thus assisted living. Consequently, many researchers have put their efforts into implementing sensor-based activity recognition systems. However, recognizing people’s natural behavior and physical activities with diverse contexts is still a challenging problem because human physical activities are often distracted by changes in their surroundings/environments. Therefore, in addition to physical activity recognition, it is also vital to model and infer the user’s context information to realize human-environment interactions in a better way. Therefore, this research paper proposes a new idea for activity recognition in-the-wild, which entails modeling and identifying detailed human contexts (such as human activities, behavioral environments, and phone states) using portable accelerometer sensors. The proposed scheme offers a detailed/fine-grained representation of natural human activities with contexts, which is crucial for modeling human-environment interactions in context-aware applications/systems effectively. The proposed idea is validated using a series of experiments, and it achieved an average balanced accuracy of 89.43%, which proves its effectiveness.

Human Action Recognition in Smart Living Services and Applications: Context Awareness, Data Availability, Personalization, and Privacy

“…Context awareness is a key aspect in designing Smart Living environments, where systems recognize, interpret, and respond to various contextual factors, including time, location, user preferences, and activities. By understanding and adapting to users' contexts, these systems can enhance user experience, promote independence, and facilitate convenience [65][66][67][68][69][70]. Some studies have focused on improving feature extraction and resolving activity confusion by using marker-based Stigmergy, a concept derived from social insects that explains their indirect communication and coordination mechanisms (Xu et al [65]).…”

“…Ehatisham-ul-Haq et al [68] propose a two-stage model for activity recognition in-the-wild (ARW) using portable accelerometer sensors. By incorporating the recognition of human contexts, the model provides a fine-grained representation of daily human activities in natural surroundings.…”

A Study on the Influence of Sensors in Frequency and Time Domains on Context Recognition

“…They evaluated model performance using six ML algorithms: Random Forest, Decision Tree, Bagging, K-Nearest Neighbor, Support Vector Machine, and Naive Bayes. The same approach was also used in [ 7 , 21 , 22 ].…”

“…Accelerometers measure linear acceleration in space; gyroscopes provide the rotation angles; and magnetometers provide the north direction relative to the device’s local reference [ 6 ]. Understanding the movement of a person is helpful in Human Activity Recognition (HAR), so recent approaches use smartphone and smartwatch accelerometers [ 7 , 8 ].…”

Human Activity Recognition by the Image Type Encoding Method of 3-Axial Sensor Data

“…Smartphones integrate various sensors such as accelerometers, gyroscopes, light sensors, and temperature sensors, making them versatile for a wide range of services such as device control and monitoring. They are also used as wearable devices for analyzing physical activity [2][3][4][5]. For this analysis, data from 3-axis accelerometers and gyroscopes are commonly used, as they provide useful information on speed, direction, and angles of human movement.…”