Human Activity Recognition Using Multi Features

Human Activity Recognition Using Multi Features

BANUSHRI S

Abstract:

Human Activity Recognition (HAR) is a multidisciplinary field that uses computational algorithms to understand and categorize human behaviors. Recently, the focus has been on utilizing multiple aspects to enhance the depth and accuracy of identification algorithms. This survey provides a comprehensive overview of the advances made in HAR, with a particular emphasis on the integration and synergistic use of various features. We explore a wide range of properties that are relevant to HAR, including spatial and temporal data, physiological indicators, and detailed environmental variables. By adopting a multi-feature approach, we can broaden the scope of recognizable activities while also improving the precision and adaptability of identification systems, addressing the dynamic and multifaceted nature of human activities. However, combining multiple attributes complicates data processing, model training, and real-time applications. This survey delves into these issues, providing insights into potential solutions and future prospects. As we enter a new era in HAR, the combination of various aspects promises a more comprehensive, context-aware, and nuanced understanding of human actions, with enormous potential for applications ranging from healthcare to smart home systems.