A camera app that can either continuously detects hand landmarks and classifies gestures

Sign language reader app

Taken from googlesample/mediapipe

Overview

This is a camera app that can either continuously detects hand landmarks and classifies gestures (gesture name and confidence level) from camera frames seen by your device’s front camera, an image, or a video from the device’s gallery using a custom task file.

The task file is downloaded by a Gradle script when you build and run the app. You don’t need to do any additional steps to download task files into the project explicitly unless you wish to use your own custom gesture recognition task. If you do use your own task file, place it into the app’s assets directory.

This application should be run on a physical Android device to take advantage of the camera.

gesture recognition demo

Build the demo using Android Studio

Prerequisites

  • The Android Studio IDE. This sample has been tested on Android Studio Dolphin.
  • A physical Android device with a minimum OS version of SDK 24 (Android 7.0 – Nougat) with developer mode enabled. The process of enabling developer mode may vary by device.

Building

  • Open Android Studio. From the Welcome screen, select Open an existing Android Studio project.
  • From the Open File or Project window that appears, navigate to and select the mediapipe/examples/gesture_recognizer/android directory. Click OK. You may be asked if you trust the project. Select Trust.
  • If it asks you to do a Gradle Sync, click OK.
  • With your Android device connected to your computer and developer mode enabled, click on the green Run arrow in Android Studio.

Models used

Downloading, extraction, and placing the models into the assets folder is managed automatically by the download.gradle file.

GitHub

View Github

Leave a Reply

Your email address will not be published. Required fields are marked *