[_O_]
SYSTEM_STANDBY
Activate the webcam module above to initialize tracking pipeline. All processing runs local-only.
/
DOWNLOADING_MODEL_ASSETS_INTO_BROWSER...Loading MediaPipe Hand Landmarker runtime (float16)HOW_IT_WORKS.log
The application analyzes the video feed using Google MediaPipe Hand Landmarker WebAssembly. The gesture is matched using Python-identical logic: Index and Middle fingers pointing upward, while Ring and Pinky fingers remain closed.
CONFIG_CONTROLSv1.0.0
BLUR_MODE
BLUR_INTENSITY40px
MIN_DETECTION_CONFIDENCE0.5
Higher values reduce false triggers but require good lighting.MAX_TRACKABLE_HANDS1
OVERLAY_RENDERS
LANDMARK_MAPPING_SPECS
GESTURE_RULE:INDEX(8)>PIP(6) & MID(12)>PIP(10)
CLOSED_RULE:RING(16)<PIP(14) & PINKY(20)<PIP(18)
ENGINE:MEDIAPIPE TASKS VISION v0.10.8
RENDERER:HTML5 CANVAS 2D CONTEXT
HARDWARE_ACCEL:GPU (WEBGL DELEGATE)
GESTURE TRACING MATCHES TARGET PYTHON SCRIPT IMPLEMENTED IN OPENCV FOR HIGH COMPATIBILITY.