A self contained example demonstrating how to use MediaPipe Hand Gesture Recognizer with Max's jweb connected to either a live webcamera stream or using still images.
This example does both handtracking and gesture recognition. If you only require handtracking try: jweb-hands-landmarker
The Gesture classification model bundle can recognize these common hand gestures:
0 - Unrecognized gesture, label: Unknown
1 - Closed fist, label: Closed_Fist
2 - Open palm, label: Open_Palm
3 - Pointing up, label: Pointing_Up
4 - Thumbs down, label: Thumb_Down
5 - Thumbs up, label: Thumb_Up
6 - Victory, label: Victory
7 - Love, label: ILoveYou
The hand landmark model bundle detects the keypoint localization of 21 hand-knuckle coordinates within the detected hand regions. The model was trained on approximately 30K real-world images, as well as several rendered synthetic hand models imposed over various backgrounds.
This example is inspired by an example by Rob Ramirez, which is in turn inspired by MediaPipe in JavaScript.