Exploring Approaches and Techniques for Human Activity Recognition in Video: A Comprehensive Overview
Preethi Salian K, Sanjeev R. Kulkarni · 2023
Recognizing human activity in videos is a difficult endeavor with many uses in monitoring, health care, sports analysis, and interaction between humans and machines. This paper delves into both traditional methods and cutting-edge deep learning-based approaches, providing a thorough overview of the advancements in the field. Through this survey, seeks to give researchers and practitioners with a comprehensive understanding of the current state-of-the-art techniques, their strengths, limitations, and future directions. Also discussed the importance and challenges associated with human activity recognition. It then presents a taxonomy of different methodologies used in this domain, including handcrafted feature-based strategies, deep learning-based methods and explores the key components involved in human activity recognition systems. Furthermore, this paper highlights the datasets commonly used for evaluating human activity recognition algorithms and provides an analysis of the performance of va rio us techniques on these datasets. In summary, this comprehensive compilation serves as a valuable resource for researchers and practitioners working in the field of human activity recognition in videos.