Application of MediaPipe Holistic pipeline that integrates integrates separate models for pose, face and hand components
Mediapipe Holistic model link: https://google.github.io/mediapipe/solutions/holistic.html
For Webcam feed change the required number in cv2.VideoCapture() For Video feed save the video feed to same folder where your jupyter notebook is hosted and pass through cv2.VideoCapture('sample_2.mp4')
Would recommend you to refer the below link on computer vision tutorials from Nick https://www.youtube.com/watch?v=pG4sUNDOZFg&list=PLXxGuqpiGy3DYadMxf1cwACoDzP0-ajcx&index=2