Classification of Machine Learning and Power Requirements for HAI (Human Activity Identification)

Human Activity Recognition(HAR) is used in many applications, such as surveillance, anti-terrorists, anti-crime securities, medical, life logging, and assistance. Besides its positive effect on well-being, the recognition of human activity has many applications. One of the key research topics in the fields of computer vision and machine learning is the human capacity for activity recognition. Pose estimation and categorization algorithms, which are now available for use on pictures or video input, it is now possible to gather and store information on several elements of human mobility in a free environment. A hierarchical structure is inherent in human activities, which can be categorized into three levels based on the nature of the action. A typical example of a movement happens when one walks, talks, stands, or sits indoors, which are everyday indoor activities. In addition, they may be more focused on activities such as those performed in kitchens or factories. However, A major challenge for activity recognition is the diversity of methods used by individuals. As a technology assistive for eldercare and healthcare, it is expected to be used mainly with the Internet of Things (IoT). The paper has demonstrated that it follows a hierarchy of analytical approaches toward the issue in a clearly defined form, and the paper showed numerous strategies that have been a part of various other studies in the field. Even though a notable amount of progress has been observed in this critical area, there is still space for further improvements in the subject topic, particularly when it comes to applying cutting-edge categorization algorithms to a variety of problems. In order to deal with these challenges, Pose estimation and classification algorithms have been used to evaluate data collection and discover human activities accurately. This study also examines the activities performed by the user during Video/Image input.