A Review on Machine Learning Techniques for Human Actions Recognition
Diana Nagpal, Rajiv Kumar · Apple Academic Press eBooks · 2023
The recognition of human activities is a remarkable research bearing in the field of computer vision. Automatic recognition of human exercises, for example, HAR has now been developed as a cutting edge zone in human–computer relationship, mobile computing, and many more. Human activity recognition provides data on a client’s conduct that permits computing frameworks to proactively help clients with their undertakings. Currently, we are using smartphone sensors to detect human activities such as accelerometer, gyroscope, barometer. In this paper, a brief understanding of human action recognition (HAR) has been provided, that is, sensor-based and vision-based HAR. The best in time techniques of machine learning such as decision trees, K-nearest neighbors have been reviewed for HAR. The results obtained by every technique and the sort of dataset they have utilized are being introduced. Also, deep learning neural network strategies have been depicted, for example, artificial neural network, recurrent neural network, and convolutional neural network, and the results obtained by these methods have also been introduced.