Voice Activity Detection in iOS

Detecting the presence of human speech in audio can be an essential component of a Speech Recognition pipeline. But whereas we humans are able to naturally distinguish speech from other sounds, machines need some help to be able to make the same distinction. Engines that do this are usually called Voice Activity Detectors or VADs for short. Given some audio input, a VAD makes a binary decision and determines whether the input contains speech or not.

Picovoice makes it simple to detect voice activity in audio data using Cobra Voice Activity Detection. It is lightweight, runs on-device and on any platform - including mobile phones. Cobra VAD performs voice activity detection locally, keeping your voice data private (i.e. it is GDPR and HIPAA-compliant by design).

Importantly, the Cobra Voice Activity Detection engine is the most accurate VAD engine across all platforms, even in comparison to Google's widely used WebRTC VAD.

In just a dozen lines of code, you can start detecting voice activity in real time from a microphone using the Cobra Voice Activity Detection iOS SDK. Let’s get started!

Install VAD SDK

The Cobra Voice Activity Detection (VAD) iOS SDK is available through CocoaPods. To add CocoaPods to a project, execute the following in your XCode project directory:

pod init

Import the Cobra-iOS binding and Picovoice's ios-voice-processor by adding the following lines to your project's Podfile:

pod 'Cobra-iOS'
pod 'ios-voice-processor'

Then, install by running:

pod install

Make sure to always open the Xcode workspace instead of the project file when building your project!

App Microphone Permissions

Add the following to the app's Info.plist file to enable recording with an iOS device's microphone:

<key>NSMicrophoneUsageDescription</key>
<string>[Permission explanation]</string>

Sign up for Picovoice Console

Sign up for a free Picovoice Console account and obtain your AccessKey. The AccessKey is only required for authentication and authorization.

Initialize Cobra in Swift

Import Cobra Voice Activity Detection and create an instance of the VAD engine with your Picovoice AccessKey:

import Cobra

do {
    let cobra = try Cobra(accessKey: "${ACCESS_KEY}")
} catch { }

Add Audio Recording

import ios_voice_processor

let audioCallback = VoiceProcessorFrameListener { frame in
    do {
        let voiceProbability = try cobra.process(pcm: frame)
        // take action based on the probability of voice
    } catch {}
}

Process Audio with Cobra

Finally, register the callback with ios-voice-processor and call .start() to begin detecting voice probability:

VoiceProcessor.instance.addFrameListener(audioCallback)

do {
    try VoiceProcessor.instance.start(
        frameLength: cobra.frameLength,
        sampleRate: cobra.sampleRate)
} catch {}

Putting it Together

All together, the example looks like this:

import Cobra
import ios_voice_processor

do {
    let cobra = try Cobra(accessKey: "${ACCESS_KEY}")

    let audioCallback = VoiceProcessorFrameListener { frame in
        do {
            let voiceProbability = try cobra.process(pcm: frame)
        } catch {}
    }

    VoiceProcessor.instance.addFrameListener(audioCallback)
    try VoiceProcessor.instance.start(
        frameLength: cobra.frameLength,
        sampleRate: cobra.sampleRate)
} catch {}

Done! It really is that easy.

For further details, visit the Cobra Voice Activity Detection product page or refer to the Cobra iOS SDK quick start guide.