DOI: 10.14704/nq.2022.20.11.NQ66323

Vision based human activity recognition using deep neural network

Kumari Priyanka Sinha, Alok Kumar, Ashish Ranjan

Abstract

Recently computer vision technology is continuously growing and the accuracy of the activity detection of human or object as well as target recognition has been improved. While working on human recognition face recognition is one of the important directions which can solve many challenges like mobile payment, safety, criminal investigation etc. Deep learning can play a vital role in the field of computer vision technology. So, in the proposed research study a neural computing model is introduced for human activity recognition and handling the other challenges. Some of the challenges like intra-class variations, image quality issues, illumination variations etc. in this research study new deep learning CNN classifiers is proposed. The experimental results shows that it achieves better result for human action recognition.

Keywords

Activity recognition, HAR, deep leaning, CNN, vision

Full Text

References