Getting Started

Licence Key

To run the DeepAR Android SDK, you first need to sign up and create your DeepAR account. If you haven't done that you can do it here: DeepAR Sign Up.

After you’ve signed up and logged in to DeepAR, you can create your new project.

Go to the Projects and click the + New Project button to create a new project.

Choose Add App for Android App.

For App Name choose any name you want, and for App ID set your bundle id.

Read more about Application ID here.

Now just hit Create app, and you will get your license key (app key).

Download the DeepAR Android SDK.

Copy the App key and paste it inside your DeepAR Android SDK project under "your_license_key_here".

DeepAR Android SDK - Installation

Android platform SDK is distributed as an aar library. Following are the steps to include the library inside your project:

• Import the deepar.aar file to your application lib folder.

• In Android Studio open Project Structure dialog (File -> Project Structure...).

• In Project Structure Dialog click "+" button and choose Import .JAR/.AAR package option then click Next.

• In the following dialog navigate to deepar.aar file in your libs directory under the File name option then click Finish.

• Close all dialogs and Sync the project. Gradle should automatically offer the option since the settings.gradle file has changed and now it should contain the following line:

    include ':deepar'

Check again Project Structure if the application module has new dependency to deepar. If not add it manually. Under dependency window click "+" -> Module dependency -> deepar -> click Ok.

Your appliciations gradle.build file should also include new entry in the dependencies section:

    implementation project(':deepar')

For a quick start on how to use the library check out the provided ExampleApp implementation.

Usage requirements

Platform

DeepAR Android SDK supports devices with Android API level 23 (Android 6.0 Marshmallow) and higher.

Hardware

There are no hard requirements, however, for best performance we recommend the following:

• 2 GB RAM

• Quad Core 1.3 GHz processor

• Launch year 2015+

Input

DeepAR SDK supports any 16:9 or 4:3 input image resolutions.

Devices must be equipped with cameras that support at least one of the mentioned resolutions.

Permissions

SDK itself doesn't check for Android permissions but certain ones are required for the library to work correctly. Following are the required permissions:

• Camera - Manifest.permission.CAMERA

• External file storage - Manifest.permission.WRITE_EXTERNAL_STORAGE

• Autofocus - android.hardware.camera.autofocus

• Record audio - android.permission.RECORD_AUDIO

It is the responsibility of user provide required permissions before using the library in the code. Within the ExampleApp code we provide the utility class PermissionActivity which implements permission check and request functionalities.

Camera feed

It is the user's responsibility to provide feed from the camera as input for the library. For this purpose, you can use the functionality of CameraGrabber class that is provided by the library. However, the class has limited functionality and may not fit your use case, therefore the source code for CameraGrabber has been provided in the example project and user can modify it to their needs.

ProGuard

If you're using ProGuard, include this rule:

    -keepclassmembers class ai.deepar.ar.DeepAR { *; }

Android 12+

Android 12 changes the policy of access to native vendor libraries. You can read about it here (see ENFORCE_NATIVE_SHARED_LIBRARY_DEPENDENCIES).

To ensure all DeepAR features work as expected (especially Background Segmentation), add the following code to your app’s manifest, inside the application tag.

AndroidManifest.xml

<uses-native-library
  android:name="libOpenCL.so"
  android:required="false" />

For Pixel devices, you may also need to add this code:

AndroidManifest.xml

<uses-native-library
  android:name="libOpenCL-pixel.so"
  android:required="false" />