an Andorid library that provide a simple API to compare the similarity between 2 faces from bitmap. This can be implemented to a face recognition system, face authentication, face matching, and many more that needed face comparison technique. This library implemented Google MLKit to detect faces and MobileFaceNet CNN model to compare faces

In your project level setting.gradle file, add the following:

Groovy DSL syntax

dependencyResolutionManagement {
		repositoriesMode.set(RepositoriesMode.FAIL_ON_PROJECT_REPOS)
		repositories {
			mavenCentral()
			maven { url 'https://jitpack.io' } // Must be added
		}
	}

Kotlin DSL syntax

dependencyResolutionManagement {
    repositoriesMode.set(RepositoriesMode.FAIL_ON_PROJECT_REPOS)
    repositories {
        google()
        mavenCentral()
        maven { url = uri("https://jitpack.io") } // Must be added
    }
}

In your app level build.gradle file, add the following dependency:

Groovy DSL syntax

dependencies {
	    implementation 'com.github.dikamahard:FaceCompareLibrary:1.0.2'
}

Koltin DSL syntax

dependencies {
	    implementation ("com.github.dikamahard:FaceCompareLibrary:1.0.2")
}

1. git clone https://github.com/dikamahard/FaceCompareLibrary
2. open your project in android studio
3. click file -> new -> import module
4. find the path of the face compare library that you cloned
5. implement the dependencies in your app level build.gradle

val fc = FaceCompare(assets)

fc.compareFaces(bitmap1, bitmap2) { result ->

    if (result) {
        // Faces match
        // Do something
        Log.d("TAG", "FACES MATCH")

    } else {
        // Faces don't match
        // Do something else
        Log.d("TAG", "FACE DO NOT MATCH")
    }

}

I know my code is messy, ineffective, and spaghetti, so feel free to make a pull requests to fix my messy code

MobileFaceNets: Efficient CNNs for Accurate Real-Time Face Verification on Mobile Devices

MobileFaceNet TensorFlow

MobileFaceNet Implementation

Google ML Kit