Using the MediaPipe library, along with OpenCV and NumPy, I created several modules involving the computer vision detection of hand and face landmarks using a camera. One module involved the use of creating a hand landmark detection class, which identified 21 points on the hand. Other projects, such as the volume control by hand gestures or finger counting, used this module to perform linear algebra operations and calculate relationships between these landmarks to perform actions such as changing the volume on the computer or counting numbers. In another project, I used MediaPipe to extract 468 landmarks (points) on the human face and then create a face mesh that was superimposed on camera feedback.
Video Demo for Finger Counting: https://www.youtube.com/watch?v=SCytboxzWVw