An Extensive Analysis of Deep Learning-Based Human Activity Detection Techniques

Because identifying human activities has several implications for a range of disciplines, including privacy, amusement, ambient community living, and healthcare oversight and control, Machine Learning (ML) and Deep Learning (DL) research is primarily focused on this area. Investigations on Human Activity Recognition (HAR) demonstrate the curiosity among investigators regarding human daily tasks. The variety of HAR system strategies are discussedin this work in an organized and understandable manner. Although the HAR discipline has advanced significantly in the last 10 years, robust hybrid ML models for the identification of actions by humans are still required. A thorough review, analysis, and open challenges of HAR systems are also done in this work. HAR systems must satisfy the application's goals, have reliable predictions, and possess a good detection rate.