AI based activity tracking and analysis system

With the increase in the automation in this world, the accuracy of the systems has increase and the cost is decreased. Due to automation the need of activity detection systems has increased which can eliminate the need of human intervention and costly infrastructure. To solve this problem deep learning is used widely. It refers to the use of neural network which classifies the various human activities. In this paper, the model used to classify the human activity is ResNet-34 which is methodology of neural network. The model has an average accuracy of 71%. The error detection and reduction algorithm is used which can increase the accuracy of prediction data considerably.