• Anchal M E
• Arpitha N
• Amrutha
• Yashmitha A
• Daya Naik, Nivin K S

This project intends human action recognition by the usage of machine learning methods, deep learning models, and computer vision techniques for analyzing and classifying human movements. Spatial features are extracted using Convolutional Neural Networks (CNNs), while dynamic patterns within human actions are extracted by Recurrent Neural Networks (RNNs). OpenCV comes in handy to implement image and video preprocess in order to extract frames efficiently and prepare data for use by deep models. To ensure high accuracy in actions for complex human actions in various scenarios, the model uses sophisticated deep learning architectures in the form of hybrid CNN-RNN. Its efficacy on bigger datasets compared to other action categories with tolerably small retraining exhibits scalability ability. The project will provide the foundation for real-time human action recognition applications in surveillance, robotics, healthcare, sports analytics, and beyond. Multimodal learning, real-time inference optimization, and integration with augmented and virtual reality platforms are to be added in future extensions.

Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs, Spatial features.

"HUMAN ACTION RECOGNITION", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.12, Issue 7, page no.667-675, July-2025, Available :http://www.jetir.org/papers/JETIRGX06127.pdf

"HUMAN ACTION RECOGNITION", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.12, Issue 7, page no. pp667-675, July-2025, Available at : http://www.jetir.org/papers/JETIRGX06127.pdf