Voice overlay helps you turn your user's voice into text, providing a polished UX while handling for you the necessary permission.

You can clone this repo, then run the Demo project by doing ./gradlew app:installDebug and launching the application:

The Voice overlay is available as a gradle dependency via JCenter. To install it, add the following line to your app's build.gradle:

dependencies {
    // [...]
    implementation 'com.algolia.instantsearch:voice:1.0.0-beta02'
    // [...]
}

1. In your Activity, check if you have the permission and show the appropriate Dialog:

if (!isRecordAudioPermissionGranted()) {
    VoicePermissionDialogFragment().show(supportFragmentManager, "DIALOG_PERMISSION")
} else {
    VoiceInputDialogFragment().show(supportFragmentManager, "DIALOG_INPUT")
}

See it implemented in the demo app.

This will display the permission dialog if the RECORD_AUDIO permission was not yet granted, then the voice input dialog once the permission is granted.

Once the user speaks, you will get their input back by implementing VoiceSpeechRecognizer.ResultsListener:

override fun onResults(possibleTexts: Array<out String>) {
    // Do something with the results, for example:
    resultView.text = possibleTexts.firstOrNull()?.capitalize()
}

See it implemented in the demo app.

If the user didn't accept the permission, you should explain the permission's rationale. If they deny the permission, you need to guide them into manually enabling it if they want to use the voice-input feature.

Voice overlay makes it easy to handle all these cases:

override fun onRequestPermissionsResult(requestCode: Int, permissions: Array<out String>, grantResults: IntArray) {
    super.onRequestPermissionsResult(requestCode, permissions, grantResults)
    if (Voice.isRecordPermissionWithResults(requestCode, grantResults)) {
        when {
            Voice.isPermissionGranted(grantResults) -> showVoiceDialog()
                shouldExplainPermission() -> showPermissionRationale(getPermissionView())
                else -> Voice.showPermissionManualInstructions(getPermissionView())
        }
    }
    // [...] eventual handling of other permissions requested by your app 
}

See it implemented in the demo app.

This will display the permission rationale when the user doesn't allow it, and the manual instructions in case they denied it.

You can customize your voice overlay in the following ways:

Several options let you adapt the voice overlay's behavior to your needs.

You can provide suggestions of what the user could say, to give them some examples.

voiceInputDialogFragment.setSuggestions(
    "64GB Smartphone",
    "Red running shoes",
    "Cheap TV screen"
)

You can prevent the overlay from automatically listening to user input.

/// Requires the user to click the mic to start listening. 
voiceInputDialogFragment.autoStart = false
// [...]
// you can also start listening programmatically with
voiceInputDialogFragment.start()

You can change any text displayed in the overlay by overriding its resource in your strings.xml:

<!-- VoiceInputDialogFragment -->
<string name="input_title_listening">Listening…</string>
<string name="input_subtitle_listening">Say something like:</string>
<string name="input_title_error">Sorry, we didn\'t quite get that.</string>
<string name="input_subtitle_error">Try repeating your request.</string>
<string name="input_hint_error">Try again</string>

<!-- VoicePermissionDialogFragment -->
<string name="permission_title">You can use voice search to find products.</string>
<string name="permission_subtitle">May we access your device’s microphone to enable voice search?</string>
<string name="permission_button_allow">Allow microphone access</string>
<string name="permission_button_reject">No</string>

<!-- Rationale/Try Again -->
<string name="permission_rationale">Voice search requires this permission.</string>
<string name="permission_button_again">Request again?</string>

<!-- Manual Instructions -->
<string name="permission_enable_rationale">Permission denied, allow it to use voice search.</string>
<string name="permission_button_enable">Allow recording</string>
<string name="permission_enable_instructions">On the next screen, tap Permissions then Microphone.</string>

You can replace the voice overlay's layouts by your own, as long as they respect the following structure:

Create a layout called voice_input.xml with

• A ViewGroup container with id @+id/voicePermission
• A View with id @+id/close for closing the overlay when clicked
• A TextView with id @+id/title
• A TextView with id @+id/subtitle

Create a layout called voice_permission.xml with

• A ViewGroup container with id @+id/voiceInput
• A VoiceMicrophone with id @+id/microphone to handle the voice input
• A TextView with id @+id/suggestions to display eventual suggestions
• A View with id @+id/close for closing the overlay when clicked
• A TextView with id @+id/title
• A TextView with id @+id/subtitle
• An eventual TextView with id @+id/hint to display a hint on error
• An eventual RippleView with id @+id/ripple if you want to keep the animation

• Need help? Ask a question to the Algolia Community or on Stack Overflow.
• Found a bug? You can open a GitHub issue.
• Questions about Algolia? You can search our FAQ in our website.

• If you want to contribute please feel free to submit pull requests.
• If you have a feature request please open an issue.
• If you use the Voice Overlay in your app, we would love to hear about it! Drop us a line on discourse or twitter.

The VoiceOverlay is available under the MIT license. See the LICENSE file for more info.