Comparative Evaluation of Human Action Recognition using Deep Learning Architectures

Recent developments in deep learning have revolutionized the field of activity recognition by humans and other entities. These advancements allow models to learn complex representations and hierarchies from raw data, hence increasing recognition accuracy. Many machine learning methods, such as histogram of gradients with k-nearest neighbor classifiers and support vector machines, have lost part of their appeal due to the extensive feature engineering and data preparation required. The aim of this work is to construct a detailed study utilizing CNN (convolutional neural networks), Inception network model, and Resnet model showcasing the promising results while removing the need for sophisticated feature manipulation. Furthermore, the study makes use of a large and diverse dataset that includes over 12,000 annotated photos illustrating various human activities. - This study advances deep learning techniques focused on human - activity recognition, showing possible improvements in efficacy and precision in real-world scenarios.