Qualcomm® AI HubAI Hub

MediaPipe-Hand-Gesture-Recognition

Real‑time hand gesture recognition optimized for mobile and edge.

The MediaPipe Gesture Recognizer is a real‑time machine learning pipeline that detects hands, predicts 21 hand landmarks, determines handedness (left/right), and classifies gestures from a predefined set

Model RepositoryHugging FaceResearch Paper

Technical Details

Input resolution:256x256
Number of parameters (PalmDetector):1.76M
Model size (PalmDetector) (w8a8):2.05 MB
Number of parameters (HandLandmarkDetector):2.72M
Model size (HandLandmarkDetector) (w8a8):3.12 MB
Number of parameters (CannedGestureClassifier):143K
Model size (CannedGestureClassifier) (w8a8):180 KB
Model size (PalmDetector) (float):6.75 MB
Model size (HandLandmarkDetector) (float):10.4 MB
Model size (CannedGestureClassifier) (float):577 KB

Applicable Scenarios

  • Gesture Control
  • Virtual Reality
  • Gaming

License

Model:APACHE-2.0

Tags

  • real-time

Supported Compute Devices

  • Snapdragon X Elite CRD
  • Snapdragon X Plus 8-Core CRD

Supported Compute Chipsets

  • Snapdragon® X Elite
  • Snapdragon® X Plus 8-Core

Related Models

See all models
View details for the MediaPipe-Face-Detection model.

MediaPipe-Face-Detection

Detect faces and locate facial features in real-time video and image streams.

View details for the MediaPipe-Pose-Estimation model.

MediaPipe-Pose-Estimation

Detect and track human face, hand, and torso in real-time images and video streams.

View details for the MediaPipe-Selfie-Segmentation model.

MediaPipe-Selfie-Segmentation

Segments the person from background in a selfie image and realtime background segmentation in video conferencing.

View details for the MediaPipe-Hand-Detection model.

MediaPipe-Hand-Detection

Real-time hand detection optimized for mobile and edge.

Looking for more? See models created by industry leaders.

Discover Model Makers