Aimybox voice assistant

Open source voice assistant built on top of Aimybox SDK

iOS version is available here

Key Features

• Provides ready to use UI components for fast building of your voice assistant app
• Modular and independent from speech-to-text, text-to-speech and NLU vendors
• Provides ready to use speech-to-text and text-to-speech implementations like Android platform speechkit, Google Cloud speechkit, Houndify or Snowboy wake word trigger
• Works with any NLU providers like Aimylogic, Rasa or Dialogflow
• Fully customizable and extendable, you can connect any other speech-to-text, text-to-speech and NLU services
• Open source under Apache 2.0, written in pure Kotlin
• Embeddable into any application or device running Android
• Voice skills logic and complexity is not limited by any restrictions
• Can interact with any local device services and local networks

How to start using

Just clone this repository and try to build and run the app module 😉

If you want some details - there is how to do the same with your hands.

1. Create a new Android project with next dependencies in the build.gradle file

    android {
        compileOptions {
            sourceCompatibility = JavaVersion.VERSION_1_8
            targetCompatibility = JavaVersion.VERSION_1_8
        }
    }

    repositories {
        jcenter()
    }
    
    dependencies {
        implementation("com.justai.aimybox:core:0.15.0")
        implementation("com.justai.aimybox:components:0.1.10")
    }

2. Add one or more dependencies of third party speech-to-text and text-to-speech libraries. For example

implementation("com.justai.aimybox:google-platform-speechkit:0.15.0")

3. Create a new project in Aimybox console, enable some voice skills and copy your project's API key.

4. Instantiate Aimybox in your Application class like that

class AimyboxApplication : Application(), AimyboxProvider {

    companion object {
        private const val AIMYBOX_API_KEY = "your Aimybox project key"
    }

    override val aimybox by lazy { createAimybox(this) }

    private fun createAimybox(context: Context): Aimybox {
        val unitId = UUID.randomUUID().toString()

        val textToSpeech = GooglePlatformTextToSpeech(context)
        val speechToText = GooglePlatformSpeechToText(context)

        val dialogApi = AimyboxDialogApi(AIMYBOX_API_KEY, unitId)

        return Aimybox(Config.create(speechToText, textToSpeech, dialogApi))
    }
}

5. Add FrameLayout to your application's layout like this

<FrameLayout
    android:id="@+id/assistant_container"
    android:layout_width="match_parent"
    android:layout_height="match_parent" />

6. Add AimyboxAssistantFragment in your activity that uses this layout (like here)

override fun onCreate(savedInstanceState: Bundle?) {
    super.onCreate(savedInstanceState)
    setContentView(R.layout.layout_activity_main)

    supportFragmentManager.beginTransaction().apply {
        replace(R.id.assistant_container, AimyboxAssistantFragment())
        commit()
    }
}

7. Make sure your app's theme contains Aimybox's styles:

<resources>
    <!-- Base application theme. -->
    <style name="AppTheme" parent="Theme.AppCompat.Light.DarkActionBar">
        <!-- Customize your theme here. -->
        <item name="colorPrimary">@color/colorPrimary</item>
        <item name="colorPrimaryDark">@color/colorPrimaryDark</item>
        <item name="colorAccent">@color/colorAccent</item>

        <!-- Customize Assistant components here -->
        <item name="aimybox_assistantButtonTheme">@style/CustomAssistantButtonTheme</item>
        <item name="aimybox_recognitionTheme">@style/CustomRecognitionWidgetTheme</item>
        <item name="aimybox_responseTheme">@style/CustomResponseWidgetTheme</item>
        <item name="aimybox_imageReplyTheme">@style/CustomImageReplyWidgetTheme</item>
        <item name="aimybox_buttonReplyTheme">@style/CustomButtonReplyWidgetTheme</item>
    </style>

    <style name="CustomAssistantButtonTheme" parent="DefaultAssistantTheme.AssistantButton">
    </style>

    <style name="CustomRecognitionWidgetTheme" parent="DefaultAssistantTheme.Widget.Recognition">
    </style>

    <style name="CustomResponseWidgetTheme" parent="DefaultAssistantTheme.Widget.Response">
    </style>

    <style name="CustomButtonReplyWidgetTheme" parent="DefaultAssistantTheme.Widget.ButtonReply">
    </style>

    <style name="CustomImageReplyWidgetTheme" parent="DefaultAssistantTheme.Widget.ImageReply">
    </style>

</resources>

Now you can run your application and tap a small microphone button in bottom right corner of the screen. Try to say some phrase that corresponds to any of enabled voice skills in your Aimybox project.

Your assistant will handle all job regarding speech recognition, processing, displaying and synthesising of the response.

More details

Please refer to the demo app to see how to use Aimybox library in your own project.

Documentation

There is a full Aimybox documentation available here. It's better to start with our Quick Start to make first steps with Aimybox.