Human Action and Face Recognition Using Python

In this paper, the goal is to develop a real-time system for human action recognition and face recognition based on advanced computer vision and machine learning technologies. It can detect human actions such as standing, sitting, bending over and doing planks by combining MediaPipe pose estimation and SimpleFacerec face recognition. It also recognizes individuals from a pre-trained facial recognition model. The proposed system can apply in fitness tracking, security systems and human-computer interaction for monitoring and analyzing the real-time behavior of humans along with their identification from videos during live streaming. It is hoped that following developments will ensure the system continues to be more accurate, refine performance and expand its range of detectable actions.