Abstract:

Abstract Human Activity Recognition has become an increasingly important research area in recent years, with applications ranging from health monitoring to human-computer interaction. This study uses the Kinetics dataset and ResNet-34 model to develop a deep learning-based method for identifying human activities. Videos of human activity are abundant in the Kinetics dataset, which we utilise to train and test our algorithm. Deep convolutional neural network called ResNet-34 has been demonstrated to be very good at classifying images. We refine our model using the Kinetics dataset to identify a range of human activities, such as sports, dancing, and regular activities. Our tests show that on the Kinetics dataset, our method performs at the cutting edge with an accuracy of above 90%. Additionally, this study compared our method to other well-known deep learning models, demonstrating that our ResNet-34 model outperforms them in terms of accuracy and efficiency. Overall, our results demonstrate the potential of deep learning-based approaches for Human Activity Recognition and provide insights for future research in this area.

Keywords: