Abstract
Human action recognition targets recognising different actions from a sequence of observations and different environmental conditions. A wide different applications is applicable to vision based action recognition research. This can include video surveillance, tracking, health care, and human–computer interaction. However, accurate and effective vision based recognition systems continue to be a big challenging area of research in the field of computer vision. This review introduces the most recent human action recognition systems and provides the advances of state-of-the-art methods. To this end, the direction of this research is sorted out from hand-crafted representation based methods including holistic and local representation methods with various sources of data, to a deep learning technology including discriminative and generative models and multi-modality based methods. Next, the most common datasets of human action recognition are presented. This review introduces several analyses, comparisons and recommendations that help to find out the direction of future research.
Highlights
Human Action Recognition (HAR) has a wide-range of potential applications
Few rules were used with the transitions of a finite state machine (FSM) to detect the fall based on the measures of the extracted bounding box
In line with the recent literature surveys for human action recognition, the most common technique used in supervised learning based models is Convolution Neural Networks (CNN)s
Summary
Human Action Recognition (HAR) has a wide-range of potential applications. Its target is to recognise the actions of a person from either sensors or visual data. Visual sensor-based methods of human action recognition are one of the most prevalent and topical areas in the computer vision research community. Observations are fed to a perception system for recognition processes These biophysical processes of the human recognition system have been investigated by many researchers to achieve similar performance in the form of computer vision systems. The structure of this review starts at low level based methods for action recognition This is followed by description of some of the important details of feature descriptor based techniques. These are transferable with respect to the performance of action recognition systems in general Thereafter, it reviews higher level feature representation based methods. The paper covers the mainstream research that has resulted in the developments of the widely known deep learning based models and their relation to action recognition systems
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have