Performance Analysis of Human Activity Recognition

Abstract

Inside the past couple of years, human action Recognition (HAR) has become an extremely imperative and significant territory of exploration because of the progression of the computerized periods of gadgets like cell phones, smartwatches, and camcorders used in our day by day lives. Profound Learning (DL) has driven analysts to utilize HAR in different spaces including wellbeing and prosperity applications the most point of this paper is to Classify the exercises happening in the casings. HAR application fields produce a major measure of information and not every one of them gives the computational force that DL models in HAR require. Utilizing our Convolutional Neural Network and Keras, we had the option to get 97.07% precision.