Human Action Recognition Using Image Preprocessing

This project utilizes ResNet50 and InceptionV3 for accurate human action classification from images, leveraging transfer learning for improved performance. Its applications span security, healthcare, smart homes, and gaming, making it a versatile solution for real-world action recognition tasks.

• The project classifies human actions from images using ResNet50 and InceptionV3 with high accuracy.

• It applies transfer learning to fine-tune pre-trained models for action recognition tasks.

• The model is evaluated with accuracy and confusion matrices for performance insights.

• It can be used in security for activity detection and monitoring.

• The system can assist in healthcare by tracking patient movements during rehabilitation.

• It can be integrated into smart homes for gesture-based control.

• The solution can enhance interactive gaming by recognizing player actions.

• It is scalable for video action recognition and robotics applications.