Human Activity Recognition

This research focuses on Human Activity Recognition (HAR) using sensor-based data, primarily from gyroscopes and accelerometers, to classify and analyze human movements in real time. With the increasing adoption of wearable devices and smart environments, HAR has gained significant importance in applications such as healthcare, fitness tracking, smart homes, and security systems. Traditional HAR methods relied on handcrafted features and classical machine learning algorithms, but recent advancements in deep learning models like Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), and Transformers have improved recognition accuracy and scalability. This study leverages the UCI HAR dataset to train and evaluate various deep learning approaches, analyzing their performance using key metrics such as accuracy, precision, recall, and F1-score. The paper also explores potential enhancements through Artificial Intelligence (AI), Augmented Reality (AR), and Virtual Reality (VR) to expand HAR capabilities in real-world applications. The insights from this research contribute to advancing HAR methodologies and paving the way for future innovations in human behavior analysis. Keywords: Human Activity Recognition, Deep Learning, Wearable Sensors, UCI HAR Dataset, CNN, RNN, Transformers